×
Log in to StudySoup
Get Full Access to Cal State Fullerton - HESC 349 - Study Guide
Join StudySoup for FREE
Get Full Access to Cal State Fullerton - HESC 349 - Study Guide

Already have an account? Login here
×
Reset your password

CAL STATE FULLERTON / Nursing and Health Science / HESC 349 / a sample of 13 small bags of the same brand of candies was selected. a

a sample of 13 small bags of the same brand of candies was selected. a

a sample of 13 small bags of the same brand of candies was selected. a

Description

School: California State University - Fullerton
Department: Nursing and Health Science
Course: Statistic
Professor: Shana charles
Term: Summer 2016
Tags:
Cost: 50
Name: HESC 349
Description: ASSIGNMENT
Uploaded: 06/07/2016
924 Pages 15 Views 8 Unlocks
Reviews


Lab Activity #1


What is the mode for ASTCUR?



Sample Experiment

Date: Feb. 5th 2016

A Simple Random Sample

Pick a simple random sample of 15 restaurants.

1. Describe the procedure.

In a simple random sample, each member of the population has an equal chance of being chosen. For example,  to get a simple random sample representing the name of restaurant, we could assign a number from one to seven to each of the city. Then, we could randomly pick a number to see which number comes up. If we pick a “one,” 

then San Jose is our sample representing the population – the name of restaurant. This is an example of simple  random sampling because every name of the restaurant has an equal chance of being selected as our sample.

2. List the 15 restaurants.

1. Emma’s Express

6. Willow Street

11. Germania

2. Stickney’s

7. Toll House

12. Poolside Grill


What is the chi-square test statistic?



3. Panda Express

8. La Fiesta

13. Le Petit Bistro

4. Santa Barb

9. Cafe Cameroon

14. Helios

5. Mariani’s

10. Garden Fresh

15. Spago’s

A Systematic Sample

Pick a systematic sample of 15 restaurants.

1. Describe the procedure. 

In a systematic sample, a starting point is selected randomly, and then every nth value of data is taken from a  listing of the population. There are 85 restaurants in total, but we want to collect a systematic sample of 15  restaurants. If we randomly choose the restaurant, El Abuelo Taq as our starting point, then we could select 15  restaurants by choosing every 6th restaurant listing.

2. Complete the table with your sample. If you want to learn more check out clinical psychology 3rd edition

1. El Abuelo Taq

4. Hung Fu

7. Lindsey’s

10. Katie’s Cafe

13. Birk’s,

2. Olive Garden


T-test or chi-square test result.



5. Tia Juana

8. Santa Barb

11. Scott’s Seafood

14. Hayes Mansion

3. Andele Taqueria

6. Pasand

9. Pacific Fresh

12. Mazeh

15. La Maison Du Cafe

A Stratified Sample

Pick a stratified sample, by the city, of 20 restaurants. Use 25% of the restaurants from each stratum.  Round to the nearest whole number. If you want to learn more check out niloufar haque

1. Describe the procedure:

In a stratified sample, the population is divided into strata (or groups), and then some members of each stratum  are randomly selected. For example, consider the population to be all restaurants in San Jose. There are 13  restaurants in San Jose, but we want to use 25% of the restaurants from each stratum (by the city). Then, we  count the total restaurants of each stratum, which is 85 in total. Thus, 13/85 multiply by 25% is 3.8 (but round  to 4) shows how many restaurants we need to use from each city.

Strata:

Restaurants in these strata:

San Jose

Emperor’s Guard, Agenda, Germania, Bamboo Hut

HESC 349 Page 1

Palo Alto

Stickney’s, Olive, Spago’s

Los Gatos

Lindsey’s, Willow Street

Mountain View

Amber Indian, La Fiesta, Le Petit Bistro

Cupertino

Samrat, Hamasushi, Helios

Sunnyvale

Pacific Fresh, Charley Brown’s, Cafe Cameroon, Faz

Santa Clara

Birk’s, Truya Sushi, Valley Plaza, Thai Pepper

2. Complete the table with your sample.

We could choose simple random samples from each of the above strata to make 20 restaurants. However, we  only want 20 random restaurants, so there is some chance to be chosen less than 4 restaurants from San Jose.   Don't forget about the age old question of which two minerals combine to form hydroxyapatite

1. Agenda

6. Helios

11. Lindsey’s

16. Stickney’s

2. Olive

7. Charley Brown’s

12. Le Petit Bistro

17. Amber Indian

3. La Fiesta

8. Truya Sushi

13. Hamasushi

18. Cafe Cameroon

4. Bamboo Hut

9. Thai Pepper

14. Pacific Fresh

19. Faz

5. Samrat

10. Emperor’s Guard

15. Willow Street

20. Birk’s

A Stratified Sample

Pick a stratified sample, by the entrée cost, of 21 restaurants. Use 25% of the restaurants from each  stratum. Round to the nearest whole number.  Don't forget about the age old question of genetics final exam study guide

1. Describe the procedure

It is same procedure I did on the previous stratified sample questions, but we now want to use the strata (the  entrée cost) and 21 restaurants. For example, 29 restaurants under $10, and the total population is 85 (all  restaurants). Thus, 29/85 multiply by 25 is 8.5 (round to 9). It all applies to each strata.

Strata:

Restaurants in these strata:

Under $10

Maharaja, Hobees, Pasand, Bamboo Hut, New Ma’s, Panda Express, Full Throttle, Thai  Pepper, Garden FreshDon't forget about the age old question of niustat

$10 to under  $15

Dawit, La Galleria, Ming’s, Emperor’s Guard, Fiesta del Mar, Charley Brown’s,  Bombay Oven

$15 to under  $20

Toll House, Blue Pheasant, Valley Plaza, Gervais, Miro’s

Over $20

Spago’s, Le Petit Bistro, Lakeside, Helios

2. Complete the table with your sample. 

1. Maharaja

6. Helios

11. Garden Fresh If you want to learn more check out rutgers anatomy

16. New Ma’s

2. Full Throttle

7. Charley Brown’s

12. Le Petit Bistro

17. Panda Express

3. Miro’s

8. Gervais

13. Lakeside

18. Valley Plaza

4. Bamboo Hut

9. Thai Pepper

14. La Galleria

19. Fiesta del Mar

5. Spago’s

10. Pasand

15. Toll House

20. Bombay Oven

21. Blue Pheasant

A Cluster Sample

Pick a cluster sample of restaurants from two cities. The number of restaurants will vary.

1. Describe your procedure.

In a cluster sample, the population is divided into groups and then some of the groups are selected randomly.  First, divide all restaurants listed on the table into two groups according to the city. Then, if we randomly  selected some of these groups, we would have a cluster sample. Our cluster sample would include all the  restaurants in those selected groups. 

Cluster Samples:

Members in this sample

San Jose

El Abuelo Taq, Pasta Mia, Emma’s Express, Bamboo Hut, Emperor’s Guard, Creekside Inn, Agenda, Gervais, Miro’s, Blake’s, Eulipia, Hayes Mansion, Germania

Santa Clara

Rangoli, Armadillo Willy’s, Thai Pepper, Pasand, Arthur’s, Katie’s Cafe, Pedro’s, La  Galleria, Birk’s, Truya Sushi, Valley Plaza, Lakeside, Mariani’s

2. Complete the table with your sample.

1. Rangoli

6. Pedro’s

11. Creekside Inn

16. Hayes Mansion

21. Emperor’s Guard

2. Valley Plaza

7. Thai Pepper

12. Pasand

17. Germania

22. El Abuelo Taq

3. Pasta Mia

8. Emma’s Express

13. Arthur’s

18. Agenda

23. Eulipia

4. Bamboo Hut

9. Birk’s

14. Katie’s Cafe

19. Armadillo Willy’s

24. Hayes Mansion

5. Lakeside

10. Truya Sushi

15. Blake’s

20. Miro’s

25. Mariani’s

Lab Activity #2

Descriptive Statistics

Date: Feb. 7th 2016

Collect the Data

Record the number of pairs of shoes you own.

1. Randomly survey 30 classmates about the number of pairs of shoes they own. Record their values.

8

11

9

10

10

7

12

6

9.5

12

6

7

7.5

9

9

10.5

6.5

9

6.5

11

11

7.5

8

7

8.5

7

8

10

11

7

3. Calculate the following values.

a. mean= 6(2)+6.5(2)+7(5)+7.5(2)+8(3)+8.5(1)+9(4)+9.5(1)+10(3)+10.5(1)+11(4)+12(2)/30        =8.7

b. 

Data

Freq.

Deviations

Deviations2

(Freq.)(Deviations2)

6

2

6­8.7 = ­2.7

(­2.7)2 = 7.29

2 x 7.29 = 14.58

6.5

2

6.5­8.7= ­2.2

(­2.2)2 = 4.84

2 x 4.84 = 9.68

7

5

7­8.7= ­1.7

(­1.7)2 = 2.89

5 x 2.89 =14.45

7.5

2

7.5­8.7= ­1.2

(­1.2)2 =1.44

2 x 1.44 = 2.88

8

3

8­8.7= ­0.7

(­0.7)2 =0.49

0.49 x 3 = 1.49

8.5

1

8.5­8.7= ­0.2

(­0.2)2 =0.04

0.04 x 1 = 0.04

9

4

9­8.7= 0.3

(0.3)2 =0.06

0.06 x 4 = 0.24

9.5

1

9.5­8.7= 0.8

(0.8)2 =0.64

0.64 x 1 =0.64

10

3

10­8.7= 1.3

(1.3)2 =1.69

1.69 x 3 = 5.07

10.5

1

10.5­8.7= 1.8 

(1.8)2 = 3.24

3.24 x 1 = 3.24

11

4

11­8.7= 2.3

(2.3)2 =5.29

5.29 x 4 = 21.16

12

2

12­8.7= 3.3

(3.3)2 =10.89

10.89 x 2 =21.78

The total is 95.25

The sample variance, s2, is equal to the sum of the last column (95.25) divided by the total number of data  values minus one (30 – 1). Therefore, 95.25/29 = 3.2845

The sample standard deviation s is equal to the square root of the sample variance. Therefore, 1.812, which is  rounded to two decimal places, s = 1.81

4. Are the data discrete or continuous? How do you know?

The sizes are continuous data since shoe size is measured.

6. Are there any potential outliers? List the value(s) that could be outliers. Use a formula to check the end  values to determine if they are potential outliers.

First, ordered from smallest to largest:

6,6,6.5,6,5,7,7,7,7,7,7.5,7.5,8,8,8,8.5,9,9,9,9,9.5,10,10,10,10.5,11,11,11,11,12,12

HESC 349 Page 1

Since there are 30 observations, the median is between the fifteenth value, 8.5, and the sixteenth value, 9. To  find the median, add the two values together and divide by two. Thus, the median is 8.75. The first quartile, Q1, is the middle value of the lower half of the data, and the third quartile, Q3, is the middle  value, or median, of the upper half of the data.

The lower half of the data are 6,6,6.5,6,5,7,7,7,7,7,7.5,7.5,8,8,8,8.5. The middle value of the lower half is 7, which is part of the data, is the first quartile. The upper half of the data

9,9,9,9,9.5,10,10,10,10.5,11,11,11,11,12,12. The middle value of the upper half is 10. Therefore, the third  quartile, Q3, is 10.

IQR = Q3 – Q1 Therefore, 10­7=3

(1.5)(IQR) = (1.5)(3) = 4.5

Q1 – (1.5)(IQR) = 10 – 4.5 = 5.5

Q3 + (1.5)(IQR) = 7 + 4.5 = 11.5

No shoe size is less than 5.5. However, 12 is more than 11.5. Therefore, 12 is a potential outlier.

Analyze the Data

1. Determine the following values.

a. Min = 6

b. M = 8.75

c. Max = 12

d. Q1 = 7

e. Q3 = 10

f. IQR = 3

5. How does the standard deviation help you to determine concentration of the data and whether or not there are potential outliers?

The standard deviation can tell us what probability of the population we can find within a certain number of  deviations.

Lab Activity #3

Running Descriptive Statistics in SPSS 

Date: Feb. 9th 2016

1. What are the numbers assigned to “La Habra” “Fullerton”, “Buena Park”, and  “Placentia” for the variable City? (Hint: Go to Variable View and look at Values for this variable).  You will need to click on the blue box that appears in the label  field to see all of the values.

Type in your answer here:  1. La Habra 2. Fullerton 3. Buena Park 4.  Placentia

2. What is variable q2 measuring? (Hint: Go to Variable View and look at Label for this variable name.  You can stretch the column by clicking on and dragging the  border of the box to see all of it).

Type in your answer here: In your neighborhood, do you have space and  equipment that are clean and safe enough for you to exercise comfortably?

3. How many variables are in this dataset?  (Hint: Go to Variable View and scroll  all the way down to the end, where you see no more variable information.  The  number of rows= the number of variables).

Type in your answer here:  12 variables

4. How many participants are in this dataset?  (Hint: Go to Data View and scroll all the way down to the end, where you see no more participants.  The number of  rows= the number of participants).

Type in your answer here:  155 participants

5. Create a table that contains the mean and standard deviation for the variable q4  (On a typical day, how many servings of fruits and vegetables do you eat?).    Hint: Go to Analyze  Descriptive Statistics  Descriptives.  Select “On a  typical day…”, click on the right arrow key to move it to the right column, and  then click OK.  Note: This mean is not an “exact mean” because the last  answer for q4 is 5+ fruits and vegetables, which does not give you an exact  number.

Copy and paste the mean and standard deviation output table here below.

Hint: To do this, you can right­click on the table, select Copy, go back to this  document, and use the Paste function to paste it here.

Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

On a typical day, how many 

servings of fruits and  vegetables do you eat?

145 1 5 2.49 .987

Valid N (listwise) 145

6. Create a frequency table for the variable q3 (In general, how would you rate  your overall health?).  Hint: Go to Analyze  Descriptive Statistics  Frequencies. Select the variable, click on the right arrow key to move it to the  right column, and then click OK.  

  

Copy and paste the frequency table here below.

In general, how would you rate your overall health?

Cumulative

Frequency Percent Valid Percent

Percent

Valid Poor 4 2.6 2.7 2.7 Fair 30 19.4 20.0 22.7 Good 58 37.4 38.7 61.3 Very Good 54 34.8 36.0 97.3 Excellent 4 2.6 2.7 100.0 Total 150 96.8 100.0

Missing System 5 3.2

Total 155 100.0

7. What is the mode for q3?  Hint: Look at the frequency table from previous  question and find the answer with the most common frequency. 

Type the answer in here: “Good” is the mode.

8. What is the median for q3?  Hint: Look at the same frequency table and find the  first answer that crosses a cumulative percent of 50.

Type the answer in here: “Good” is the median.

9. Create a bar chart for the variable q6_1 (Choosing lower calorie drinks).  Hint: Go to Analyze  Descriptive Statistics  Frequencies. Select the variable, click  on the right arrow key to move it to the right column.  Then go to Charts (far right hand side), select Bar Chart, select Continue, and then click OK.  

Copy and paste the bar chart here below.

Lab Activity #4

Probability

Date: Feb. 18th 2016

Table 3.11 Population

Color

Quantity

Yellow (Y)

6

Green (G)

7

Blue (BL)

5

Brown (B)

9

Orange (O)

6

Red (R)

7

Table 3.12 Theoretical Probabilities

With Replacement

Without Replacement

P(2 reds)

7/40x7/40=0.031

7/40x6/39 =0.027

P(R1B2 OR B1R2)

(7/40x9/40)+(9/40x7/40)= 0.078

(7/40x9/39)+(9/40x7/39)=0.08

P(R1 AND G2)

7/40x7/40=0.0306

7/40x7/39=0.0314

P(G2IR1)

49/98=1/2, 0.5

7/39=0.1795

P(no yellows)

34/40x34/40=0.7225

34/40x33/39=0.71923

Table 3.13 Empirical Results

With Replacement Without Replacement

(R1,B2) (G1,O2)

(Y1,R2) (O1,G2)

(Y1,R2) (B1,Y2)

(BL1,B2) (R1,G2)

(BL1,G2) (R1,O2)

(G1,R2) (BL1,Y2)

(Y1,R2) (G1,B2)

(O1,BL1) (G1.R2)

(O1,BL2) (R1,G2)

(R1,O2) (Y1,R2)

(G1,R2) (B1,BL2)

(BL1,B2) (O1,R2)

(R1,Y2) (G1,R2)

(R1,Y2) (B1,BL2)

(B1,G2) (BL1,O2)

(O1,Y2) (BL1,R2)

(Y1,B2) (G1,R2)

(B1,R2) (G1,B2)

(B1,BL2) (R1,O2)

(Y1,G2) (BL1,R2)

(Y1,G2) (BL1,B2)

(G1,R2) (Y1,R2)

(B1,R2) (G1,BL2)

(R1,G2) (Y1,BL2)

Table 3.14 Empirical Probabilities

With Replacement

Without Replacement

P(2 reds)

0/24+0/24=0

0/24+0/23=0

P(R1B2 OR B1R2)

1/24+1/24=0.08333

0/24+1/23=0.043478

P(R1 AND G2)

5/24+4/24=0.375

4/24+4/23=0.340579

P(G2IR1)

1/24=0.04167

2/24=0.08333

P(no yellows)

19/24=0.79167

16/24=0.6666

Discussion Questions

1. Replacement leaves open the possibility that the individual is randomly chosen a  second time. When we sample with replacement, the two sample values are independent.  Practically, this means that what we get on the first one does not affect what we get on  the second. In sampling without replacement, the two sample values are not independent,  which means what we get on the first one affects what we can get for the second one.

2. a. Theoretical “With Replacement”: P(no yellows) = 0.7225

b. Empirical “With Replacement”: P(no yellows) = 0.7917

    c. The decimal values of both theoretical and empirical result are close. I expected  them not to be that much closer because yellow M&Ms could be either picked up or not;  thus, the value of empirical probabilities would totally depend on the random chance of  all possible outcomes.

3. If the total number of observed occurrences is increased to 240 times, the number of  ways the specific event occurs will be increased as well. Therefore, the value of the  probability that an even will occur is going to be changed too. This is because that  empirical probability is an event that the event will happen based on how often the event  occurs after running an experiment.

4. As an experiment is repeated more and more times, the proportion of outcomes  favorable to any particular even would tend to come closer to the theoretical probability  of that event due to the same probability of the outcomes.

5. There is different value of sample space (the set of all possible outcomes) between  P(G1 AND R2) and P(R1IG2). For P(G1 AND R2), if picking two M&Ms out of 40 one  at a time, there are 40(40) = 1600 outcomes, the size of the sample space. However, for  P(R1IG2), the sample space would be reduced to those outcomes that already have a red  on the first M&M. There are 49 + 49 = 98 possible outcomes. 

Assignment #5  

Date: Feb. 25th 2016

4.1 Probability Distribution Function (PDF) for a Discrete Random Variable 

69. Suppose that the PDF for the number of years it takes to earn a Bachelor  of Science (B.S.) degree is given in the following table.

x

P(x)

3

0.05

4

0.40

5

0.30

6

0.15

7

0.10

a. In words, define the random variable X.

X = the number of years taken to complete a B.S.

b. What does it mean that the values zero, one, and two are not included for  x in the PDF?

The probability is 0 that someone completes the degree in less than 3 years.

4.2 Mean or Expected Value and Standard Deviation 

70. A theater group holds a fund-raiser. It sells 100 raffle tickets for $5  apiece. Suppose you purchase four tickets. The prize is two passes to a  Broadway show, worth a total of $150.

a. In words, define the random variable X.

X = net profit

b. List the values that X may take on.

S = (­20, 130)

c. Construct a PDF.

The probability of winning is simply the proportion of tickets you hold. P(X = 130) = 0.04          P(X = ­20) = 0.96

d. If this fund-raiser is repeated often and you always purchase four tickets,  what would be your expected average winnings per raffle?

Expected average= (0.04)(130) + (0.96)(­20) = ­14

71. A game involves selecting a card from a regular 52-card deck and tossing a coin. The coin is a fair coin and is equally likely to land on heads or tails. • If the card is a face card, and the coin lands on Heads, you win $6

• If the card is a face card, and the coin lands on Tails, you win $2 • If the card is not a face card, you lose $2, no matter what the coin shows.

The variable of interest is X = net gain or loss, in dollars.

The face cards are J, Q, K (Jack, Queen, King). There are (3)(4) = 12 face cards and 52 – 12 = 40 cards that are not face cards.

We first need to construct the probability distribution for X. We use the card and coin  events to determine the probability for each outcome, but we use the monetary value of X  to determine the expected value.

Card Event

$X net gain or loss

P(X)

Face Card and Heads

6

(12/52)(1/2) = 6/52

Face Card and Tails

2

(12/52)(1/2) = 6/52

(Not Face Card) and (H or T)

–2

(40/52)(1) = 40/52

a. Find the expected value for this game (expected net gain or loss). Expected value = (6)(6/52) + (2)(6/52) + (–2) (40/52) = –32/52, = –$0.62

b. Explain what your calculations indicate about your long-term average  profits and losses on this game.

If I play this game repeatedly, over a long number of games, I would expect to lose 62 cents  per game, on average.

c. Should you play this game to win money?

No, I should not play this game to win money because the expected value indicates an  expected average loss.

72. You buy a lottery ticket to a lottery that costs $10 per ticket. There are  only 100 tickets available to be sold in this lottery. In this lottery there are  one $500 prize, two $100 prizes, and four $25 prizes. Find your expected  gain or loss.

(Gain­Loss) x probability = expected value

$500 prize: +490 x 1/100 = 4.90

$100 prize: +90 x 2/100   = 1.80

$ 25 prize: + 15 x 4/100   = 0.60

no prize: ­10 x 93/100   = ­9.30

Therefore, 4.90+1.80+0.60­9.30= ­2, which I expected loss.

4.3 Binomial Distribution 

Use the following information to answer the next four exercises. Recently, a  nurse commented that when a patient calls the medical advice line claiming  to have the flu, the chance that he or she truly has the flu (and not just a  nasty cold) is only about 4%. Of the next 25 patients calling in claiming to  have the flu, we are interested in how many actually have the flu.

83. Define the random variable and list its possible values.

X = the number of patients calling in claiming to have the flu, who actually have the flu. X  = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 

84. State the distribution of X.

X ~ B(25, 0.04) 

85. Find the probability that at least four of the 25 patients actually have the  flu.

P(x>4) = binomcdf(25, 0.04, 4)

The given probability would be 0.0165 

86. On average, for every 25 patients calling in, how many do you expect to  have the flu?

25 x 0.04 = 1

Only one.

5.1 Continuous Probability Functions 

72. Consider the following experiment. You are one of 100 people enlisted to  take part in a study to determine the percent of nurses in America with an  R.N. (registered nurse) degree. You ask nurses if they have an R.N. degree.  The nurses answer “yes” or “no.” You then calculate the percentage of  nurses with an R.N. degree. You give that percentage to your supervisor.

a. What part of the experiment will yield discrete data?

If random variable, X, is equal to the number of people with an R.N, then X is a discrete  random variable. So, it yields discrete data.

b. What part of the experiment will yield continuous data?

If X is anyone who has an R.N. degree, then measure its value by answering “yes” or “no,”  X is a continuous data. So, it yields continuous date.

73. When age is rounded to the nearest year, do the data stay continuous, or do they become discrete? Why?

The data stays continuous because age is measured, not counted.

HESC 349, Spring 2016

Assignment #6  

Mar 03 2016

For each of the following questions, please put your answers in this template  below (feel free to expand the sections to show your work). Submit the  completed document by Thursday of Week 6 at midnight.

Chapter 7

7.1 The Central Limit Theorem for Sample Means (Averages) Yoonie is a personnel manager in a large corporation. Each month she must  review 25 of the employees. From past experience, she has found that the  reviews take her approximately 5 hours each to do with a population  standard deviation of 1.2 hours. Let Χ be the random variable representing  the time it takes her to complete one review. Assume Χ is normally  distributed. Let X ¯ be the random variable representing the mean time to  complete the 25 reviews. Assume that the 25 reviews represent a random  set of reviews.

1. What is the mean, standard deviation, and sample size? Mean = 5 hours

Standard Deviation = 1.2 hours

Sample size = 25

2. Complete the distributions.

a. X ~  N( 5, 1.2)

b. �̅~  N(5, 1.2/√25)

3. Find the probability that one review will take Yoonie from 4.5 to 5.25  hours. Sketch the graph, labeling and scaling the horizontal axis. Shade the  region corresponding to the probability.

Since μ=5 and  =1.2 σ

We need to compute P (4.5 ≤ X ≤ 5.25). 

The corresponding z­values needed to be computed are:

Therefore, we get:

a.

(Shading area represents the probability, which is 0.2441)

b. P (4.5 < x < 5.25) = 0.2441

By using calculator, normalcdf(4.5, 5.25, 5, 1.2) = 0.24405 

4. Find the probability that the mean of a month’s reviews will take Yoonie  from 4.5 to 5.25 hrs. Sketch the graph, labeling and scaling the horizontal  axis.  

a. Shade the region corresponding to the probability.

    (Shading area represents the probability, which is 0.8326)

b. P (4.5, 5.25) = 0.8326

When calculating the standard deviation for the mean using the Central Limit Theorem in  this problem, the sample size is n = 25, the number of reviews she completes in a month. By using calculator, normalcdf(4.5, 5.25, 5, 1.2/√25)=0.8326

5. What causes the probabilities in Exercise 7.3 and Exercise 7.4 to be  different?

The fact that the two distributions are different accounts for the different  probabilities.

6. Find the 95th percentile for the mean time to complete one month's  reviews.  

a. Sketch the graph.

Let K = the 95th percentile. Find K, where P(�̅  < 0.95)

K = 5.39

b. The 95th Percentile = 5.39 hours

P(4.5<�<̅5.25)= invNorm(0.95, 5, 1.2/√25) = 5.3947

HESC 349, Spring 2016

Assignment #7  

Mar 10 2016

For each of the following questions, please put your answers in this template  below (feel free to expand the sections to show your work). Submit the  completed document by Thursday of Week 7 at midnight.

Chapter 8

8.4 | Confidence Interval (Home Costs) 

Collect the Data 

Check www.zillow.com. Record the sale prices for 35 randomly selected  homes recently

listed in Orange county in the table below. Use this information for the rest of the Assignment.

1. Complete the table:

Table 1: Home Prices in Orange County, March 2016

$400,000

$650,000

$499,000

$898,000

$450,000

$349,900

$828,000

$777,000

$565,000

$485.000

$530,990

$454,990

$525,000

$900,000

$559,000

$439,000

$499,900

$405,990

$360,000

$489,990

$679,000

$405,000

$759,000

$425,000

$1490,000

$850,000

$500,000

$500,000

$775,000

$399,900

$699,900

$455,000

$400,000

$349,900

$650,000

Source: www.zillow.com

Describe the Data 

1. Compute the following:

    _

a. x  = 582,985

b. sx = 227651 (from the calculator)

c. n = 35

 _

2. In words, define the random variable X.

X bar represents the average sale price of a sample.

3. State the estimated distribution of the data. Use both words and symbols.

We should use the Student‐T because the standard deviation of the population is  unknown.

Use the Data to Construct Confidence Intervals 

1. Using the given information, complete Table 2 and construct a confidence  interval for each confidence level given.

Table 2: Confidence Intervals for Home Prices in Orange County, March 2016

Confidence Level

EBM/Error Bound

Confidence Interval

50%

26243.4

(556742, 609228)

80%

52371.3

(530614, 635356)

95%

78191.4

(504749, 661176)

99%

104973.6

(478011, 687959)

Source: www.zillow.com and student’s analysis

50%

df = 35 – 1 = 34 CL so α = 1 – CL = 1 – 0.5 = 0.5

α/2 = 0.25,  tα/2 = t0.25

The area to the right of t0.25 is 0.25, and the area to the left of t0.25 is 1 – 0.25 = 0.75 Tα/2 = t0.25 = 0.682 using invT(.75,34) on the calculator

EBM =(tα/2)(s/√n) 

          =0.682 x (227651/√35) = 26243.4

x ˉ– EBM = 582,985 – 26243 = 556742

x ˉ+ EBM = 582,985 + 26243 = 609228

The 95% confidence interval is (556742, 609228)

We estimate with 95% confidence that the home price in Orange County is between 556742,  609228

80%

df = 35 – 1 = 34 CL so α = 1 – CL = 1 – 0.8 = 0.2

α/2 = 0.1,  tα/2 = t0.1

The area to the right of t0.1 is 0.1, and the area to the left of t0.1 is 1 – 0.1 = 0.9 Tα/2 = t0.1 = 1.361 using invT(.9,34) on the calculator

EBM =(tα/2)(s/√n) 

          =1.361 x (227651/√35) = 52371.3

x ˉ– EBM = 582,985 – 52371 = 530614

x ˉ+ EBM = 582,985 + 52371 = 635356

The 95% confidence interval is (530614, 635356)

We estimate with 80% confidence that the home price in Orange County is between 530614,  635356

95%

df = 35 – 1 = 34 CL so α = 1 – CL = 1 – 0.95 = 0.05

α/2 = 0.025,  tα/2 = t0.025

The area to the right of t0.025 is 0.025, and the area to the left of t0.025 is 1 – 0.025 = 0.975

Tα/2 = t0.025 = 2.032 using invT(.975,34) on the calculator.

EBM =(tα/2)(s/√n) 

          =2.032 x (227651/√35) = 78191.4

x ˉ– EBM = 582,985 – 78191 = 504749

x ˉ+ EBM = 582,985 + 78191 = 661176

The 95% confidence interval is (504749, 661176)

We estimate with 95% confidence that the home price in Orange County is between 504749,  661176

99%

df = 35 – 1 = 34 CL so α = 1 – CL = 1 – 0.99 = 0.01

α/2 = 0.005,  tα/2 = t0.005

The area to the right of t0.005 is 0.005, and the area to the left of t0.005 is 1 – 0.005 = 0.995 Tα/2 = t0.0005 = 2.728 using invT(.995,34) on the calculator

EBM =(tα/2)(s/√n) 

          =2.728 x (227651/√35) = 104973.6

x ˉ– EBM = 582,985 – 104974 = 478011

x ˉ+ EBM = 582,985 + 104974 = 687959

The 95% confidence interval is (478011, 687959)

We estimate with 99% confidence that the home price in Orange County is between 478011,  687959

2. What happens to the EBM as the confidence level increases? Does the  width of the confidence interval increase or decrease? Explain why this  happens.

As the confidence level increases, the EBM is accordingly increased. Then, the width of the confidence interval is increased. Due to the larger level of confidence level, it makes  sense that the larger level of confidence interval is wider. To be more confident that the  confidence interval actually does contain the true value of the population mean of home  prices, the confidence interval necessarily needs to be wider. 

HES 349

ASSIGNMENT #8

MAR 17TH 2016

Background of the SPSS dataset 

This data is the Adolescent survey portion of the 2014 California Health Interview  Survey. It contains answers regarding a wide variety of demographics, health status, and  health behaviors questions from interviews with teenagers (ages 12­17) from January  2014 to December 2014.

The actual questions (what you need to answer and turn in)

1. Create a frequency table for the variable ASTCUR.   Hint: Go to Analyze  Descriptive Statistics  Frequencies.  Select “Current Asthma”, click on the right arrow key to move it to the right column, and then click OK.

Copy and paste the frequency output table here below.

CURRENT ASTHMA

Cumulative

Frequency Percent Valid Percent

Percent

Valid NO CURRENT ASTHMA 915 87.0 87.0 87.0 CURRENT ASTHMA 137 13.0 13.0 100.0 Total 1052 100.0 100.0

2. What is the mode for ASTCUR?  Hint: Look at the frequency table from  previous question and find the answer with the greatest frequency. 

Type the answer in here: “No Current Asthma” is the mode.

3. Create a bar chart for the variable HGHTI_P (height:inches).  Hint: Go to  Analyze  Descriptive Statistics  Frequencies. Select the variable, click on the  right arrow key to move it to the right column.  Then go to Charts (far right hand  side), select Histogram, select Continue, and then click OK.  

Copy and paste the histogram here below.

4. What is the mean for HGHTI_P?  Hint: Look at the histogram output from the  previous question. 

Type the answer in here: The mean is 64.44inches.

5. Run a one sample t­test for ASTCUR to test if the CHIS sample is significantly  different from 0, which would mean no asthma in the population. Hint: Go to  Analyze  Compare Means  One­Sample T Test. Select the variable, click on  the right arrow key to move it to the right column.  Then click OK.  

Copy and paste the t­test output table here below.

One­Sample Statistics

N Mean Std. Deviation Std. Error Mean

CURRENT ASTHMA 1052 .13 .337 .010

One­Sample Test

Test Value = 0

Mean

95% Confidence Interval of the Difference

Lower Upper

CURRENT 

t df Sig. (2­tailed)

Difference

ASTHMA12.544 1051 .000 .130 .11 .15

6. Does the CHIS sample population have a significantly higher rate than 0 of  asthma? How do you know?

Type the answer in here: Since the Sig (p­value) is less than 0.05 (the alpha), we  reject the null; the sample is different than the CHIS sample population that has a significantly higher rate than 0 of asthma,

7. Run a one sample t­test for HGHTI_P to test if the CHIS sample is significantly  different from 68 (the average height in the U.S.). Hint: Go to Analyze  Compare Means  One­Sample T Test. Select the variable, click on the right  arrow key to move it to the right column.  Change the test value to 68. Then click  OK.  

Copy and paste the t­test output table here below.

One­Sample Statistics

N Mean Std. Deviation Std. Error Mean

HEIGHT ­ INCHES (PUF 

RECODE)1052 64.44 4.550 .140

One­Sample Test

Test Value = 68

95% Confidence Interval of

the Difference

Lower Upper

HEIGHT ­ INCHES (PUF 

t df Sig. (2­tailed)

Mean

Difference

RECODE)­25.381 1051 .000 ­3.561 ­3.84 ­3.29

8. Does the CHIS sample population have a significantly different average height  than 68 inches tall? How do you know?

Type the answer in here: Since Sig. (p­value) is less than 0.05, we reject the null;  the sample is different than the CHIS sample population that has a significantly  different average height than 68 inches tall. 

9. Run a one sample t­test for HGHTI_P to test if the CHIS sample is significantly  different from 64 (a value very close to the mean of HGHTI_P). Hint: Go to  Analyze  Compare Means  One­Sample T Test. Select the variable, click on  the right arrow key to move it to the right column.  Change the test value to 64.  Then click OK.  

Copy and paste the t­test output table here below.

One­Sample Statistics

N Mean Std. Deviation Std. Error Mean

HEIGHT ­ INCHES (PUF 

RECODE)1052 64.44 4.550 .140

One­Sample Test

Test Value = 64

95% Confidence Interval of the

Difference

Lower Upper

HEIGHT ­ INCHES (PUF

t df Sig. (2­tailed)

Mean

Difference

RECODE)3.130 1051 .002 .439 .16 .71

10. Does the CHIS sample population have a significantly different average height  than 64 inches tall? How does sample size affect whether the difference is  significant?

Type the answer in here: Since Sig. (p­value) is less than 0.05, we reject the null; the  sample is different than the CHIS sample population that has a significantly different  average height than 64 inches tall. Higher sample size allows to increase the  significance level of the findings, since the confidence of the result are likely to  increase with a higher sample size. This is to be expected because larger the sample  size, the more accurately it is expected to mirror the behavior of the whole group.  Therefore if we want to reject the null hypothesis, then you should make sure your  sample size is at least equal to the sample size needed for the statistical significance  chosen and expected effects.

Assignment #9

For this assignment, use the 2014 Adolescent California Health Interview  Survey as the dataset, and do your calculations in SPSS. Input your answers  in the space provided below.

Research Question: Is there a significant difference between the rates (i.e. the  means) of having current asthma between girls and boys in the 2014 CHIS  Adolescent sample?

1. H0: The rates of having current asthma between girls and boys are equal.

Mean of Girls = Mean of Boys

2. HA: The rates of having current asthma between girls and boys are significantly different.  

Mean of Girls ≠ Mean of Boys

3. In words, define the random variable that is being tested (i.e. compared  for two different groups).

X= Self-Reported Gender Group (Female, Male)

4. The distribution to use for the test is the normal distribution. 5. Determine the test statistic using your data.  

Hint: Go to Analyze  Compare Means  Independent Samples T-test. Select  “Current Asthma”, and click on the right arrow key to move it to the right  column in the “test variable” box. Scroll down to “Gender,” and click on the  right arrow key to move it to the right column in the “grouping variable” box.  Click on “define groups” and enter 1 for group 1 and 2 for group 2, and then  click “continue.” Then click “OK” to run the test.

Copy and paste the t-test output table here.

Page 1 of 2

Group Statistics

SELF­REPORTED GENDER N Mean Std. Deviation Std. Error Mean CURRENT ASTHMA FEMALE 494 .12 .322 .014 MALE 558 .14 .349 .015

Independent Samples Test

Levene's Test

for Equality of

Variances t­test for Equality of Means

95% Confidence

Interval of the

Difference

Lower Upper

CURRENT

F Sig. t df

Equal 

Sig.

(2­

tailed )

Mean

Differen ce

Std. Error Difference

ASTHMA

variances 

assumed

Equal 

variances not  assumed

5.445 .020 ­1.162 1050 .245 ­.024 .021 ­.065 .017 ­1.168 1048.120 .243 ­.024 .021 ­.065 .016

Hint: To do this, you can right-click on the table, select Copy, go back to this  document, and use the Paste function to paste it here.

6. Determine the p-value using the SPSS table.

Type the answer here. The p-value is 0.020

7. Do you or do you not reject the null hypothesis? Why?

Type the answer here. Due to the p-value (less than 0.05), the null  hypothesis would be rejected.  

8. Write a clear, substantively significant conclusion using complete  sentences.

Type the answer here. According to the statistical results, we could see the p-value of “equal variances assumed” is less than 0.05. Because of  that, the null hypothesis that the rates of current asthma between girls  and boys are same would be rejected. Therefore, we might judge the  result to be statistically significant. However, we should notice the  variance of boys’ group has more population than girls – 64. When  

Page 2 of 2

considering the difference of their sample size, we could otherwise say  that the 0.024 – the value of mean differences between girls and boys – is  approximately zero value. Indeed, the resulting difference score of two  variable groups is close to zero; therefore, we could conclude that there is not much difference between girls and boys. By examining representative  samples, we could produce more accurate effect size estimates; for  examples, a population.  

Page 3 of 2

Assignment #10

For this assignment, use the 2014 California Health Interview Survey as the  dataset, and do your calculations in SPSS. Input your answers in the space  provided below.

Research Question: Is there a significant relationship between getting a flu shot and  gender?

1. H0: Two factors (getting a flu shot and gender) are independent.  There is no significant relationship between getting a flu shot and  gender.  

2. HA: Two factors (getting a flu shot and gender) are associated  together, not independent (dependent). There is significant  relationship between getting a flu shot and gender.

3. Run the cross-tab of the two variables and the chi-square test. Show the  contingency table results for comparing these two variables.

Hint: Go to Analyze  Descriptive Statistics  Crosstabs. Select “Flu Shot”,  and click on the right arrow key to move it to the right column in the “rows”  box. Scroll down to “Self-reported Gender,” and click on the right arrow key to move it to the right column in the “columns” box. Click on “statistics” and  check the box for chi-square, and then click “continue.” Click on “cells” and  check the box for “column percentage,” and then click “continue.” Then click  “OK” to run the test.

Copy and paste the cross-tab output table here.

Hint: To do this, you can right-click on the table, select Copy, go back to this  document, and use the Paste function to paste it here.

HAD FLU SHOT/FLUMIST IN PAST 12 MOS * SELF­REPORTED GENDER Crosstabulation

SELF­REPORTED GENDER

MALE

271

48.6%

287

51.4%

FEMALE Total

HAD FLU SHOT/FLUMIST  IN PAST 12 MOS

NO Count 252 523

% within SELF­REPORTED 

GENDER51.0% 49.7% YES Count 242 529

% within SELF­REPORTED 

GENDER49.0% 50.3%

Total Count 494 558 1052 Page 1 of 2

% within SELF­REPORTED 

GENDER100.0% 100.0% 100.0%

4. What is the chi-square test statistic?  

Copy and paste the chi-square table output table (just below the crosstab  table on the output) here.

Hint: To do this, you can right-click on the table, select Copy, go back to this  document, and use the Paste function to paste it here.

Type the answer here.

Chi­Square Tests

Asymptotic

Value df

Significance (2­ sided)

Exact Sig. (2­ sided)

Exact Sig. (1­ sided)

Pearson Chi­Square .627a 1 .428

Continuity Correctionb.533 1 .465

Likelihood Ratio .627 1 .428

Fisher's Exact Test .459 .233 Linear­by­Linear Association .626 1 .429

N of Valid Cases 1052

a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 245.59. b. Computed only for a 2x2 table

5. What is the p-value of the chi-square test statistic?

Type the answer here. The p-value of the chi-square is 0.428

6. Do you reject or fail to reject the null hypothesis? Why?

Type the answer here. Due to the p-value of the chi-square(more than  0.05), we would fail to reject the null hypothesis.

7. Write a clear, substantive conclusion using a complete sentence.

Type the answer here. Since no α is given, assume α = 0.05. P value = 0.428, α ≺ p-value. Therefore, since α≺ p-value, we would  fail to reject the null hypothesis, which means the two factors –  getting a flu shot and gender are independent. At a 5% level of  significance, from the data, there is sufficient evidence to conclude  that there is a no significant relationship between getting a flu shot  and gender.

Page 2 of 2

Page 3 of 2

Assignment #11

For this assignment, use the 2014 California Health Interview  Survey Adolescent dataset, and do your calculations in SPSS. Input  your answers in the space provided below. Delete the italicized  instructions and hints, but keep the questions in your submission.

Research Question: Does height in inches have a significant association with  weight for adolescents?

1. Write the null and alternative hypotheses for testing height’s effect on  weight.

Type the answers in complete sentences here. The null hypothesis  indicates whether there is no a significant linear  

relationship(association) between weight in pounds and height in  inches for adolescents; the alternative hypothesis would be there is  a significant linear relationship between weight and height.

2. Which variable is the dependent variable (that is, the one being affected)?  Which variable is the main independent variable of interest?

Type the answers in complete sentences here. The dependent  variable is weigh whereas the main independent is height in inches.

3. Show the bivariate (or “simple”) linear regression results for describing the association between height in inches and weight in pounds.

Hint: Go to Analyze  Regression  Linear. Select “Weight - pounds”,  and click on the right arrow key to move it to the right column in the  “Dependent” box. Scroll to “Height - inches,” and click on the right  arrow key to move it to the right column in the “Independent(s)” box.  Click on “statistics” and check the box for confidence intervals, and  then click “continue.” Then click “OK” to run the test.

Copy and paste the bivariate linear regression output tables here.

Model Summary

Model R R Square Adjusted R Square Std. Error of the Estimate 1 .618a.382 .381 26.418 a. Predictors: (Constant), HEIGHT ­ INCHES (PUF RECODE)

Page 1 of 3

Unstandardized Coefficients

Coefficientsa Standardized

Coefficients

95.0% Confidence Interval for B

Model

B Std. Error Beta

t Sig.

Lower Bound

Upper Bound

1 (Constant) ­162.834 11.568 ­14.076 .000 ­185.533 ­140.135

HEIGHT ­ INCHES 

(PUF RECODE)4.560 .179 .618 25.462 .000 4.208 4.911 a. Dependent Variable: WEIGHT ­ POUNDS (PUF RECODE)

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

4. What is the value of the coefficient for height in inches in the bivariate  linear regression model? What is the p-value for height in inches?

Type the answers in complete sentences here. The value of the  coefficient of height is inches is 4.560, and the p-value is 0.00.

5. Show the mutivariate linear regression results for describing the  association between height in inches and weight in the past year, controlling  for the following additional independent variables: 1) # of cans of soda and  2) # of times eating fast food.

Hint: Go to Analyze  Regression  Linear. Select “Weight-pounds”,  and click on the right arrow key to move it to the right column in the  “Dependent” box. Scroll to “Height-inches,” and click on the right  arrow key to move it to the right column in the “Independent(s)” box.  Do the same with the other two independent variables, adding them all to the “Independent(s)” box. Click on “statistics” and check the box for confidence intervals, and then click “continue.” Then click “OK” to run  the test.

Copy and paste the multivariate linear regression output tables here.

Model Summary

Page 2 of 3

Model R R Square Adjusted R Square Std. Error of the Estimate 1 .619a.383 .381 26.417 a. Predictors: (Constant), # TIMES ATE FAST FOOD DURING PAST WEEK (PUF 1 YR RECODE), HEIGHT ­  INCHES (PUF RECODE), # CANS OF SODA W/SUGAR DRUNK YESTERDAY (PUF 1 YR RECODE)

Coefficientsa 

Unstandardized Coefficients

Standardized Coefficients

95.0% Confidence Interval for B

Model

B Std. Error Beta

t Sig.

Lower Bound

Upper Bound

1 (Constant) ­163.519 11.577 ­14.124 .000 ­186.237 ­140.801

HEIGHT ­ INCHES 

(PUF RECODE)4.549 .179 .616 25.381 .000 4.197 4.901 # CANS OF SODA 

W/SUGAR DRUNK  YESTERDAY (PUF 1 YR RECODE)

# TIMES ATE FAST  FOOD DURING  PAST WEEK (PUF 1  YR RECODE)

.492 1.158 .010 .425 .671 ­1.780 2.764 .835 .651 .032 1.282 .200 ­.443 2.113

a. Dependent Variable: WEIGHT ­ POUNDS (PUF RECODE)

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

6. What are the p-values of each of the three independent variables in the  multivariate linear regression model?  

Type the answers in complete sentences here. The p-value for  height in inches is 0.00, the p-value of # cans of soda is 0.671, and  the p-value of # times ate fast food is 0.200.

7. Which, if any, of the independent variables have a significant relationship  with weight in pounds in the multivariate model?  

Page 3 of 3

Type the answers in complete sentences here. Only the height in  inches has a significant relationship with weight due to the p-value,  which is less than 0.05. Other two independent variables have  higher p-value than 0.05.

8. Does the significance of height in inches change from the bivariate to the  multivariate model?

Type the answer in complete sentences here. Yes it does. We would  notice that the r-square value has increased from 0.382 to 0.383 but very small. And, the bivariate coefficient value has also changed  from 4.560 to 4.549.  

9. From the results of the multivariate linear regression, do you reject or do  you fail to reject the null hypothesis in question 1? Why or why not?

Type the answer in complete sentences here. I would reject the null  hypothesis because the p-value, 0.00, is less than the significance  level of 0.05.

10. Write a clear conclusion regarding the association between height in  inches and weight in pounds, using the multivariate model results.

Type the answer in complete sentences here. The p-value of the  independent variable of interest shows whether the association  between height in inches and weight in pounds. From the  multivariate model results, we would notice that the p-value is 0.00, which is less than 0.05. Therefore, we could conclude from the p value and reject the null hypothesis that there is an association  between height in inches and weight in pounds for adolescents.  Also, the r-squared value, 0.383, is the percent of variation in the  dependent variable that can be explained by variation in the  independent variable whereas 0.617 is the percent of variation in  weights that is not explained by variation in height. Therefore,  there is a significant linear relationship between height and weight  because the correlation coefficient is significantly different from  zero.

Page 4 of 3

Assignment #12

For this assignment, use the 2014 California Health Interview Survey  Adolescent, and do your calculations in SPSS. Input your answers in the  space provided below. Delete the italicized instructions and hints, but keep  the questions in your submission.

Research Question: Is there a significant association between height in  inches and visiting an emergency room (ER) in the past year?

1. Write the null and alternative hypotheses for testing height’s association  with ER visit in the past year.

Type the answers in complete sentences here. The null hypothesis (H0)  indicates whether there is no a significant association between  emergency room in the past year (X1) and height in inches (X2); the  alternative hypothesis (HA) would be there is a significant  relationship between emergency room in the past year (X1) and  height in inches (X2).

2. Which variable is the dependent variable (that is, the one being affected)?  Which variable is the main independent variable of interest?

Type the answers in complete sentences here. The dependent  variable is emergency room in the past year, and the independent  variable is height in inches.

3. Show the bivariate (or “simple”) logistic regression results for describing  the association between height in inches and visting and ER in the past year, including the odds ratios.

Hint: Go to Analyze  Regression  Binary logistic. Select “ER Visit”,  and click on the right arrow key to move it to the right column in the  “Dependent” box. Scroll to “Height in inches,” and click on the right  

arrow key to move it to the right column in the “Covariates” box. Click  on “save” and check the box for probabilities, and then click  “continue.” Then click “OK” to run the test.

Copy and paste the final logistic regression output table here.

Model Summary

Step ­2 Log likelihood Cox & Snell R Square Nagelkerke R Square Page 1 of 3

1 1061.466a.004 .006 a. Estimation terminated at iteration number 4 because parameter estimates changed by less than .001.

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

Step 1a HEIGHT ­ INCHES (PUF 

RECODE).034 .017 3.959 1 .047 1.035 Constant ­3.557 1.111 10.257 1 .001 .029 a. Variable(s) entered on step 1: HEIGHT ­ INCHES (PUF RECODE).

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

4. What is the value of the odds ratio for height in inches in the bivariate  logistic regression model? What is the p-value for height in inches?

Type the answers in complete sentences here. The odds ratio for  height in inches is 1.035, and the p-value is 0.047.

5. Show the mutivariate logistic regression results for describing the  association between height in inches and having an ER visit in the past year,  controlling for the following additional independent variables: 1) family size,  and 2) weight in lbs, including the odds ratios.

Hint: Go to Analyze  Regression  Binary Logistic. Select “ER Visits”,  and click on the right arrow key to move it to the right column in the  “Dependent” box. Scroll to “Height in inches,” and click on the right  arrow key to move it to the right column in the “Covariates” box. Do  the same with the other two independent variables, adding them both  to the “Covariates” box. Click on “save” and check the box for  probabilities, and then click “continue.” Then click “OK” to run the test.

Copy and paste the final multivariate logistic regression output table  here.

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

Page 2 of 3

Model Summary

Step ­2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 1052.892a.012 .019 a. Estimation terminated at iteration number 4 because parameter estimates changed by less than .001.

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

Step 1a HGHTI_P ­.002 .021 .007 1 .933 .998 FAMSIZE2_P1 ­.073 .064 1.322 1 .250 .929 WGHTP_P .007 .003 7.200 1 .007 1.007 Constant ­1.932 1.243 2.417 1 .120 .145 a. Variable(s) entered on step 1: HGHTI_P, FAMSIZE2_P1, WGHTP_P.

6. What are the p-values of the independent variables in the multivariate  logistic regression model?  

Type the answers in complete sentences here. The p-value for  family size is 0.250, the p-value for weight in pounds is 0.007, and  the p-value of height in inches is 0.933.

7. Which, if any, of the independent variables have a significant relationship  with having an ER visit in the multivariate model?  

Type the answers in complete sentences here. Due to the p-value  that is less than 0.05, the independent variable of weight in pounds  has a significant relationship with having an ER visit; otherwise, two other independent variables – height and family size – have the p value that is more than 0.05.

8. How does the significance of height in inches change from the bivariate to  the multivariate model?

Type the answer in complete sentences here. There are some  changes of the significance of height in inches from bivariate to the  multivariate; first of all, the p-value has been changed from 0.047  to 0.933, the r-square from 0.004 to 0.012, the odds ratio from  1.035 to 0.998, and the B coefficient from 0.034 to -0.002.

Page 3 of 3

9. From the results of the multivariate logistic regression, do you reject or do  you fail to reject the null hypothesis in question 1? Why or why not?

Type the answer in complete sentences here. From the results of  the multivariate regression, I would fail to reject the null hypothesis due to the p-value that is more than 0.05.

10. Write a clear conclusion regarding the association between height in  inches and having an ER visit in the past year, using the multivariate model  results.

Type the answer in complete sentences here. As we have learned  that the p-value of an independent variable (X2=height in inches) in  a logistic regression equation shows the significance of the  association with the dependent variable (X1=ER visit), holding all  other variable constant. By given the result of the p-value; that is,  0.933, we would conclude to fail to reject the null hypothesis, which indicates that there is no statistically significant association  between emergency room in the past year (X1) and height in inches  (X2). This is because that a larger (insignificant) p-value suggests that changes in the predictor are not associated with changes in the response.

Page 4 of 3

Assignment #13

For this assignment, use the 2014 California Health Interview Survey  Adolescent, and do your calculations in SPSS. Input your answers in the  space provided below. Delete the italicized instructions and hints, but keep  the questions in your submission. Use the One-Way ANOVA procedure in this  hypothesis test, treating the variable “number of times eating fast food” as a categorical variable.

Research Question: Is there a significant association between family size (a  ratio variable) and number of times eating fast food in the past week (an  ordinal variable)?

1. Write the null and alternative hypotheses for testing the differences in  mean family size in each group of number of times eating fast food in the  past week.

Type the answers in complete sentences here. The null hypothesis  (H0) is that there are no differences in mean of family size (µ1) and  number of times eating fast food (µ2) in the past week. The  alternative hypothesis (H1) is that there are statistically different  means between family size (µ1) and number of times eating fast  food (µ2) in the past week.

2. Show the one-way ANOVA results for mean family size in each group of  number of times eating fast food in the past week.

Hint: Go to Analyze  Compare Means  One-Way ANOVA. Select  “family size”, and click on the right arrow key to move it to the right  column in the “Dependent List” box. Scroll to “# of times eating fast  food,” and click on the right arrow key to move it to the right column in the “Factor” box. Click on “options” and check the box for Descriptive,  and then click “continue.” Click on “post hoc” and check the box for  Tukey, and then click “continue.” Then click “OK” to run the test.

Copy and paste the “Oneway” tables here.

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

Page 1 of 2

Descriptives

FAMILY SIZE: INCL ALL PEOPLE SUPPORTED BY HH INCOME (PUF 1 YR RECODE)   95% Confidence Interval for

Mean

N Mean Std. Deviation Std. Error

Lower Bound Upper Bound Minimum Maximum

0 TIMES 274 3.94 1.188 .072 3.80 4.08 2 8 1 TIME 368 3.81 1.117 .058 3.70 3.93 2 7 2 TIMES 235 3.97 1.317 .086 3.81 4.14 2 8 3 TIMES 94 4.35 1.389 .143 4.07 4.64 2 8 4 TIMES 44 4.27 1.128 .170 3.93 4.62 3 8 5+ TIMES 37 4.57 1.591 .261 4.04 5.10 3 8 Total 1052 3.98 1.239 .038 3.90 4.05 2 8

ANOVA

FAMILY SIZE: INCL ALL PEOPLE SUPPORTED BY HH INCOME (PUF 1 YR RECODE)   Sum of Squares df Mean Square F Sig.

Between Groups 40.208 5 8.042 5.343 .000 Within Groups 1574.198 1046 1.505

Total 1614.406 1051

3. What is the value of the F statistic? What is the p-value of the F statistic?

Type the answers in complete sentences here. The F statistic value  is 5.343 whereas the p-value of the F statistic is 0.00.

4. Do you reject or fail to reject the null hypothesis, based on the F test and  p-value?

Type the answers in complete sentences here. Based on the F test  and p-value, I would reject the null hypothesis.

5. Show the post hoc Tukey results from the one-way ANOVA mean family  size in each group of number of times eating fast food in the past week.

Page 2 of 2

Hint: Go to Analyze  Compare Means  One-Way ANOVA. Select  “family size”, and click on the right arrow key to move it to the right  column in the “Dependent List” box. Scroll to “# of times eating fast  food,” and click on the right arrow key to move it to the right column in the “Factor” box. Click on “options” and check the box for Descriptive,  and then click “continue.” Click on “post hoc” and check the box for  Tukey, and then click “continue.” Then click “OK” to run the test.

Copy and paste the “Post Hoc Tests” table here.

Hint: To do this, you can right-click on the table in SPSS, select Copy  Special, select “.jpg” format to copy, go back to this document, and  use the Paste function to paste it here.

Multiple Comparisons

Dependent Variable:   FAMILY SIZE: INCL ALL PEOPLE SUPPORTED BY HH INCOME (PUF 1 YR RECODE)   Tukey HSD   

(I) # TIMES ATE FAST  FOOD DURING PAST  WEEK (PUF 1 YR  RECODE)

(J) # TIMES ATE FAST FOOD DURING PAST  WEEK (PUF 1 YR  RECODE)

Mean

Difference (I J)

95% Confidence Interval

Std.

Error

.098

.109

.147

.199

.215

Sig.

.775

1.000

.059

.557

.042

Lower

Bound

­.15

­.34

­.83

­.90

­1.24

.098

.102

.142

.196

.212

.775

.611

.002

.175

.005

­.41

­.45

­.94

­1.02

­1.36

.109

.102

.150

.202

.217

1.000

.611

.120

.677

.070

­.28

­.13

­.80

­.87

­1.21

Upper

Bound

0 TIMES 1 TIME .129 .41 2 TIMES ­.033 .28 3 TIMES ­.409 .01 4 TIMES ­.331 .24 5+ TIMES ­.626*­.01

1 TIME 0 TIMES ­.129 .15 2 TIMES ­.162 .13 3 TIMES ­.539*­.13 4 TIMES ­.460 .10 5+ TIMES ­.755*­.15

2 TIMES 0 TIMES .033 .34 1 TIME .162 .45 3 TIMES ­.377 .05 4 TIMES ­.298 .28 5+ TIMES ­.593 .03

3 TIMES 0 TIMES .409 .147 .059 ­.01 .83 1 TIME .539*.142 .002 .13 .94 2 TIMES .377 .150 .120 ­.05 .80

Page 3 of 2

4 TIMES .078 .224 .999 ­.56 .72

5+ TIMES ­.217 .238 .944 ­.90 .46

.199

.196

.202

.224

.274

.557

.175

.677

.999

.890

­.24

­.10

­.28

­.72

­1.08

4 TIMES 0 TIMES .331 .90 1 TIME .460 1.02 2 TIMES .298 .87 3 TIMES ­.078 .56 5+ TIMES ­.295 .49

5+ TIMES 0 TIMES .626*.215 .042 .01 1.24 1 TIME .755*.212 .005 .15 1.36 2 TIMES .593 .217 .070 ­.03 1.21 3 TIMES .217 .238 .944 ­.46 .90 4 TIMES .295 .274 .890 ­.49 1.08 *. The mean difference is significant at the 0.05 level.

6. Using the Post Hoc test results, which, if any, of the categories in number  of times eating fast food have significant differences from each other?  

Type the answers in complete sentences here. By given the Post  Hoc test results, there are six of the categories in number of times  eating fast food that have significant differences from each other;  the group of 0 times compared to more than 5 times, the group of 1  time compared to 3 and more than 5 times, the group of 3 times  compared to 1 time, and the group of more than 5 times compared  to 0 and 1 time.

7. Write a clear conclusion regarding whether there are any differences in  mean family size among the categories of number of times eating fast food.

Type the answer in complete sentences here. By given the results of ANOVA and Post-Hoc test, we could notice when comparing the  differences between all possible of individual groups, the ANOVA  test told us that there is at least one group is significantly different  from at least one other group. Indeed, we have found from the  above question; there are six different categories in the number of  times eating fast food that have significant differences from each  other. Furthermore, the p-value of the F statistic, which was 0.00, is  less than the alpha value of 0.05; therefore, there is strong  evidence that we would reject the null hypothesis and conclude that there is statistically different means between family size (µ1) and  number of times eating fast food (µ2) in the past week.

Page 4 of 2

Peer Reviewed

Title:

The Recession Index: Measuring the Effects of the Great Recession on Health Insurance Rates and Uninsured Populations

Journal Issue:

California Journal of Politics and Policy, 7(2) 

Author:

Charles, Shana Alex, UCLA Center for Health Policy Research

Snyder, Sophie, UCLA Center for Health Policy Research

Publication Date:

2015

Permalink:

http://escholarship.org/uc/item/8r42w28w 

DOI:

http://dx.doi.org/10.5070/P2cjpp7225516 

Acknowledgements:

The authors gratefully acknowledge the initial contributions to the formulation of this study by the late E. Richard Brown. We also thank the programming staff at the Center for their invaluable support.

Author Bio:

Director of Health Insurance Studies, UCLA Center for Health Policy Research and Assistant Researcher, Department of Health Policy and Management

Graduate student researcher, UCLA Center for Health Policy Research and PhD student, Department of Health Policy and Management

Keywords:

recession impact, health insurance, uninsured, geographic variation

Local Identifier:

 cjpp_25516

Abstract:

<html/>

Copyright Information:

All rights reserved unless otherwise indicated. Contact the author or original publisher for any necessary permissions. eScholarship is not the copyright owner for deposited works. Learn more at http://www.escholarship.org/help_copyright.html#reuse

eScholarship provides open access, scholarly publishing services to the University of California and delivers a dynamic research platform to scholars worldwide.

The Recession Index: Measuring the Effects of the Great Recession  on Health Insurance Rates and Uninsured Populations 

Shana Alex Charles, Sophie Snyder 

UCLA

Abstract

Background: The economic recession that began in 2008, or the “Great Recession,” did not  affect all counties in California equally. With differential effects of economic indicators such as  job loss, it is possible that differential effects were also seen in health insurance rates by county  and demographic group. Objective: To study whether the Great Recession had a differential im

pact on the uninsured rates among counties in California. Research Design: A four-level “reces sion index” measured the impact of various economic indicators, between the populations unin sured for all or part of the prior year in 2009 compared to 2007. Methods: Data sources include  the 2007 and 2009 California Health Interview Surveys and California Employment Develop ment Department unemployment data. Results: The medium recession impact group (that is,  counties with high increases in unemployment and lower household incomes on average) had the  highest growth in the uninsured rates, due to a large drop in job-based coverage only partially  offset by public coverage. Changes in coverage by demographic groups were similar among re cession index categories. Conclusions: We find that the uninsured in 2009 were older, more  likely to be U.S.- born citizens, had lower household incomes, and were more likely to be unem ployed and looking for work, regardless of the impact of the Great Recession at the county level.  The growth in the uninsured rates in the medium-impact group highlights the importance of pub lic health insurance programs as a safety net during economic downturns.

1

The Recession Index: Measuring the Effects of the Great Recession  on Health Insurance Rates and Uninsured Populations 

Shana Alex Charles, Sophie Snyder 

UCLA

Introduction

Beginning in 2008, California experienced an economic recession of greater proportions than  the rest of the United States. Unemployment in the state was consistently at least two percentage  points higher than the rest of the country. In mid-2007, California’s unemployment rate stood at  5.5 percent. By mid-2009, the unemployment rate had soared to 12.3 percent (CA EDD 2010).  

Corresponding with this increase in unemployment, the population uninsured for all or part  of the past year swelled from 6.4 million in 2007 to 7.1 million in 2009 (Lavarreda and Snyder  2012). The effects of the recession were widespread, enduring, and, for the most part, detri mental to access to health care. An investigation of the health insurance consequences of the re cession exposes potential shortfalls in the health care system.

The existing research literature documents the Great Recession’s contribution to substantial  losses or changes in health insurance coverage at the national level. Variations in insurance cov erage at the regional level have not been examined, (Greenstein and Sherman 2008) and no re search has been done that examines the relationships between regional economic indicators (i.e., unemployment and personal income), public health insurance programs, and health insurance  status. The study of these relationships is essential to understanding the role that public coverage  programs play throughout California’s diverse regions.  

The safety net was clearly crucial during the recession but was not sufficient to avoid cover age gaps. The decline in employment-based insurance among adults during the Great Recession  was only partially offset by Medicaid expansions (Rowland 2009). The aggregate increase in  Medicaid enrollment caused spending on this public program to increase 14.5 percent to $387  billion between 2007 and 2009 (Holahan et al. 2011; Young et al. 2013).

Adults accounted for nearly all of the increase in the number of uninsured. Fifty million non elderly Americans had no health insurance in 2009, up 5.6 million from 2007 (Fronstin 2010;  Holahan et al. 2011). Many children gained coverage through Medicaid, thereby harboring this  age group from joining the uninsured ranks in large numbers.

While Medicaid enrollment grew during the recession, employment-based health insurance  (EBHI) declined substantially. EBHI is the largest source of insurance for individuals below age  65 (Christianson et al. 2011; Gould 2005; Gould 2012; Holahan et al. 2011; Levit et al. 2013;  Roehr 2010). From 2007 to 2009, unemployment in the US rose dramatically (5.1 percentage  points) (US Bureau of Labor Statistics 2010). Consequently, much of the shift in health coverage  was due to heightened unemployment during the recession.  

2

The loss of health insurance that accompanied much of the rising unemployment compound ed other stress factors brought about by the recession such as loss of income. Median household  income declined 4.2 percent and the number of people living in poverty increased 16.9 percent  between 2007 and 2009 (US Bureau of Labor Statistics 2010). Falling income is noteworthy be cause many health care costs are paid out of pocket.  

Many employers that did not drop health coverage shifted the cost to employees (Christian son et al. 2011). This cost shifting strained disposable income during and after the Great Reces sion and further reduced access to health care (Gilmer and Kronick 2009). The associations be tween income, access to health care, and health outcomes are well established in the literature  

(Doty et al. 2009; Freeman et al. 1987; Shi et al. 1999). These relationships highlight the detri mental health toll related to the economic aspects of the Great Recession.

California was no exception to the rule, but the regional consequences of the Great Recession  are less well documented in the literature than the national effects. At 12.3 percent, California  had one of the highest jobless rates in the nation in 2009, up from 5.5 percent in 2007 (CA EDD  2012). During the recession, Californians living in middle-income families fell to less than 50  percent, and the gap between the highest and lowest income families expanded (Bohn and Schiff 2011). The toll on income was longer-term in California than in the US at large. In California,

median household income increased slightly between 2007 and 2009 (0.7%) but fell 4.25 percent  between 2007 and 2010 (US Bureau of Labor Statistics 2010).  

Policymakers in California reacted unfavorably to the recession in light of rising unemploy ment and falling income, restricting enrollment in safety-net programs such as Medi-Cal due to  mounting budget deficits (Redlener and Grant 2009). In 2008, the state increased the frequency  with which parents and children were required to renew coverage, retracting the guarantee of  full-year coverage essential for children with ongoing medical needs. In addition, California re quired children and parents on Medi-Cal to comply with new reporting procedures that may have  caused sizeable gaps in coverage for many families (Ross and Marks 2009).  

Some variation in outcomes was related to unemployment rates. Sacramento’s unemploy ment rate was higher than San Francisco’s by October 2009 (US Bureau of Labor Statistics  2010). Some variation among regions was the result of specific policy decisions. Los Angeles  County’s substantial budget reductions ($140 million for the 2008–09 fiscal year) led to cuts in  Medi-Cal program administration and mental health services. San Diego County’s reduced pay ments to Medi-Cal providers limited access to services (Redlener and Grant 2009).

This study examines the relationship between economic changes heightened by the Great Re cession—namely unemployment and loss of household income—and the loss of health insurance. We seek to answer the following questions: (1) did the increased impact of the Great Recession  at county-level lead to increased loss of health insurance for households in those counties, (2)  which demographic groups were most impacted by the loss of health insurance, and (3) did the difference vary by the impact of the Great Recession where a person lived?

Methods

Our study examines whether economic variations in regions throughout California are asso ciated with effects of the Great Recession on health insurance coverage. We use the 2007 and  2009 California Health Interview Surveys (CHIS) and link data from adult, child, and adolescent  interviews. We use data from CHIS and the California Employment Development Department  (EDD) to create a “recession index” by county that will shed light on health insurance coverage  

3

trends in the regions. The index captures variations in unemployment and decreases in household  income at the county level, based on the expanded Andersen Model of social determinants of  health care access that includes environmental-level (i.e., societal) variables (Andersen 1995).

We incorporated six county-level measures into the recession index, representing the nearest  government level that affected the health system in the geographic area: (1) 2009 unemployment  rate higher than state average; (2) the increase in the unemployment rate from 2007 to 2009  higher than the state average; (3) 2009 mean household income lower than state average; (4)  whether there was a decrease in mean household income from 2007 to 2009; (5) 2009 median  household income lower than state average; and (6) whether there was a decrease in median  household income from 2007 to 2009.  

Factors were chosen to measure the economic environment in which a person could access  health services. Both the mean and median income measures were included as proxies to evalu ate the income inequality in a county in the absence of a GINI index. We then divided the coun ties into four groups based on the number of factors that were either true for the average of their  sampled population in the CHIS (household income indicators) or for the county as a whole ac cording to the EDD data.

“Low impact” counties had either 0 or 1 of the recession index indicators. “Moderate impact” counties had 2 or 3 indicators. “Medium impact” counties had 4 indicators, and “High impact” counties had 5 or 6 indicators. In CHIS, counties have been aggregated for sampling purposes  into strata, with the smaller counties grouped together. No stratum contained less than 400 un weighted cases in CHIS, and the strata cannot be disaggregated into the counties that make up a  group. Therefore, counties that have been grouped into a single stratum were given the same  number for their aggregated recession index score. There are 44 total county-level strata in the  CHIS survey. In the final groupings of the recession index, 15 counties fell into the low-impact  category, 10 were in each of the moderate impact and medium impact categories, and 9 were in  the category of highest recession impact (Table 1).

The CHIS is the largest state-based health survey in the nation, with over 40,000 household  interviews for each sample of the survey. The survey is administered in six languages (English,  Spanish, Mandarin, Cantonese, Vietnamese, and Korean) to give a full picture of California’s  multiethnic/multilingual population. The survey sample is stratified into 44 counties and county  groupings (for some rural counties with small populations) to ensure representative samples for  most counties. County stratification enables policy analysis that takes into account the variations  across counties, most notably county differences in unemployment rates and in county-based  public health care insurance programs for children. Using CHIS 2007 and 2009 data enables us  to examine the impact of the economic recession that began in the final quarter of 2008.

“Uninsured” is defined as any period of one month or greater of not having medical insur ance coverage within the 12 months prior to the CHIS interview. At the household level, we in clude household income as a percentage of the Federal Poverty Level (FPL) in all models, which  takes into account the income of both the adult CHIS respondent and the spouse (if applicable). At the individual level, we included work status of the adult CHIS respondent and citizen and  immigration status of all respondents.

We used bivariate and logistic regression models, with chi-squared to determine significant  differences among cells in the bivariate analyses. Odds ratios and p-values determined signifi cant differences in the logistic regression model, with uninsured status as the dependent variable.

4

Table 1. Counties by County-Level Recession Index, Ages 0–64, California, 2009

County-Level  

Recession  

Index

Low  

Impact

Moderate  

Impact

Medium  

Impact

High  

Impact

County

Alameda 

Contra Costa 

El Dorado 

Marin 

Napa 

Orange 

Placer 

San Diego 

San Francisco 

San Luis Obispo 

San Mateo 

Santa Barbara 

Santa Clara 

Santa Cruz 

Sonoma

Alpine 

Amador 

Calaveras 

Inyo 

Los Angeles 

Mariposa 

Mendocino 

Mono 

Nevada 

Riverside 

Sacramento 

Shasta 

Stanislaus 

Tuolumne 

Ventura 

Yolo

Butte 

Del Norte 

Humboldt 

Lake 

Lassen 

Madera 

Modoc 

Monterey 

Plumas 

Sierra 

Siskiyou 

Solano 

Trinity 

Tulare 

Yuba

San Bernardino 

Alpine 

Calaveras 

Imperial 

Plumas 

San Joaquin 

Shasta 

Trinity 

Yuba

Sources: 2007 and 2009 California Health Interview Surveys and California EDD unemployment data.

Results

Discrepancies in the growth of the uninsured populations among the four county groups were  evident. While the three lower impact groups all had some increase in the rate of uninsured (and  a statistically significant increase for the medium impact group), the highest impact group actual ly saw a slight drop in the percentage of uninsured (Table 2). All four groups experienced at least  some decline in the rate of employment-based coverage. For counties in the lowest two catego ries of recession impact, these declines were smaller, and for the moderate impact group, they  were not significantly different statistically from 2007. The medium impact group experienced  the highest drop in job-based health insurance from 2007 to 2009.

The increase in public coverage (either Medi-Cal or Healthy Families) is the other major fac tor in the differences in uninsurance (Table 3). High recession impact counties expanded their  public coverage programs, while the rest of the county groups had smaller increases or none at  all. These expansions offset some of the losses in insurance coverage in these counties.  

Although the size of the uninsured populations in each county group differed, the patterns of  changing demographics among them were markedly similar. In all four of the county groups, the  uninsured populations from 2007 to 2009 shifted slightly towards an older population, with  growth in those ages 45–64 years in three county groups and growth in ages 26–44 years in the  lowest recession impact group. This is consistent with the loss of employment among middle

class, middle-aged workers during this time period, although it remained true that younger work ers were less likely to have coverage in the first place.

5

Table 2. Insurance Status by County-Level Recession Index Group, Ages 0–64, California,  2007 and 2009  

Insurance Status

Uninsured

Employment-Based  Health Insurance  (EBHI)

Medi-Cal Health  

Insurance

Rate

95 percent  CI

Rate

95 percent  

CI

Rate

95 percent  

CI

Recession Level Impact Scale

Low  

Recession  Impact

2007

19.1

(18.3-19.9)

56.9

(56.0-57.8)

13.9

(13.3-14.6)

2009

20.8

(19.6-19.9)

53.6

(52.2-57.8)

14.2

(13.4-14.6)

Moderate  Recession  Impact

2007

17.1

(15.2-19.1)

58.1

(55.6-60.6)

15.9

(13.9-17.9)

2009

18.3

(15.8-19.1)

57.4

(53.9-60.6)

15.0

(12.6-17.9)

Medium  

Recession  Impact act

2007

20.8

(19.0-22.6)

52.2

(50.1-54.3)

18.0

(16.5-19.6)

2009

26.2

(22.8-22.6)

45.1

(41.5-54.3)

18.8

(16.3-19.6)

High  

Recession  Impact

2007

22.5

(20.2-24.8)

47.9

(45.4-50.3)

22.4

(20.3-24.6)

2009

21.5

(18.8-24.8)

43.8

(40.2-50.3)

25.8

(22.6-24.6)

Sources: 2007 and 2009 California Health Interview Surveys, CA EDD unemployment data.

In each county group, the uninsured population shifted from 2007 to 2009 to a lower overall  household income, with a greater proportion of the uninsured in 2009 having income at or below  the income level that would be included under the future ACA Medi-Cal expansion.5 The coun ties with the highest recession impact experienced a slight (but not statistically significant) in crease in the percentage of uninsured having incomes over 400 percent FPL, from 10 percent in  2007 to 12.7 percent in 2009.

For all but the highest impact recession group, the proportion of US-born or naturalized citi zens among the uninsured grew from 2007 to 2009. Ranging from an increase of 4 percentage  points in the low impact group to 7.9 percentage points in the medium impact group, this trend  clearly shows how the composition of the uninsured population changed with the loss of job based coverage during the recession. High recession impact counties are mainly counties in  which the noncitizen populations (with and without “green cards”) comprise a substantial pro portion of the residents.  

The work status of the uninsured population underwent the most dramatic shift from 2007 to  2009 as the state absorbed an increase in the unemployment rate of 6.8 percentage points (more  than doubling the 2007 rate). This dramatic shift occurred in every county group, with drops in  the proportion of the uninsured who were working full-time ranging from 14.5 percentage points  (low impact) to 17.1 percentage points (high impact). While some of these newly uninsured  dropped entirely out of the work force, the largest increases were among the uninsured individu

als who were unemployed and looking for work. Small increases were seen in the proportion of  6

Table 3. County-Level Recession Index Group by Demographics, Ages 0–64, California,  2007 and 2009

Recession Level  Impact Scale

Low Recession  Impact

Moderate  

Recession Impact

Medium  

Recession Impact

High Recession  Impact

2007

2009

2007

2009

2007

2009

2007

2009

Age Group

0 to 18 Years

15.4

13.8

17.2

14.4

20.0

19.2

16.2

13.6

19 to 25 Years

22.0

19.6

18.4

19.7

23.9

24.0

23.7

21.1

26 to 44 Years

39.2

43.2

42.0

41.6

35.4

33.3

42.4

41.9

45 to 64 Years

23.4

23.4

22.4

24.4

20.7

23.5

17.8

23.4

Total

100%

100%

100%

100%

100%

100%

100%

100%

Federal Poverty Level

0-133 percent  

FPL

37.3

42.0

44.3

47.9

38.2

41.1

50.0

53.6

134-400 percent  FPL

42.0

37.2

41.7

39.7

48.4

45.6

40.0

33.8

401%+ FPL

20.8

20.8

14.0

12.4

13.3

13.4

10.0

12.7

Total

100%

100%

100%

100%

100%

100%

100%

100%

Work Status

Full-Time

63.3

48.8

61.9

46.0

64.4

47.9

64.7

47.6

Part-Time

11.2

10.6

8.1

13.1

9.0

11.9

7.7

9.3

Employed, Not  at Work

---

---

---

---

---

---

---

---

Unemployed,  

Looking for  

Work

8.1

18.2

8.8

21.2

6.6

21.9

9.1

20.5

Unemployed,  

Not Looking for  Work

16.4

21.6

20.4

18.7

19.3

18.0

18.2

20.3

Total

100%

100%

100%

100%

100%

100%

100%

100%

Citizenship and Immigration Status

US Born or Nat uralized Citizen

67.4

71.4

62.7

67.5

71.6

79.5

71.8

66.3

Noncitizen with  Green Card

13.1

14.4

16.3

14.2

13.3

10.4

12.1

15.3

Noncitizen  

without Green  Card

19.5

14.2

21.0

18.3

15.1

10.1

16.1

18.4

Total

100%

100%

100%

100%

100%

100%

100%

100%

Sources: 2007 and 2009 California Health Interview Surveys, CA EDD unemployment data. Data are unstable due to coefficient of variation above 30%.

* Numbers are rates and will not add to 100%.

7

uninsured individuals who had part-time employment, particularly in the moderate and medium  recession impact categories.

When controlling for all demographic factors in a multivariate model, only the medium im pact recession group exhibited an increased likelihood of being uninsured from 2007 to 2009  (OR = 1.39, p < 0.001; Table 4). The other two groups, moderate and high impact, had no statis tically significant difference compared to the reference group, which contained the low recession  impact counties. Expected demographic indicators showed significant differences in all county  groups. Notably, those who were unemployed and not looking for work did not show a signifi cant difference in their odds of being uninsured compared to full-time employees.

Discussion

This article presents data on the disparate impact of the economic recession on county  groups within California and the health insurance types and coverage status of their residents. We found that counties with the highest impact of the recession did not have the highest rate of  growth in the uninsured population. This was mainly because of the smaller decline in an already  low rate of job-based coverage and an increase in the rate of public coverage, likely due to the  very low household incomes of the uninsured in the highest-recession impact group. Additional ly, the medium-recession impact group, which did have the largest increase in uninsurance, in cludes many of the smaller, more rural counties in the state that have less outreach and enroll ment infrastructure for public coverage assistance.

The decline in employer-based insurance was offset by increases in public coverage, demon strating the importance of public programs during economic downturns. Public coverage eligibil ity during this time depended heavily on age, with children’s eligibility for CHIP and Medicaid  far exceeding that of their parents, and childless able-bodied adults having no eligibility at all.  Our results show that children were much less likely to be uninsured than any other age group,  supporting the positive impact that public coverage eligibility had for this population. We also  found that the uninsured population overall, with some variation among county groups, changed  in composition. In 2009, the uninsured population was older, had lower household incomes, and  was more likely to have US citizenship and less likely to have full-time employment compared  to the 2007 uninsured population.  

This shift has policy implications for the implementation of both Medi-Cal expansion and  Covered California (the California Health Benefit Exchange) under the Affordable Care Act of  2010 (ACA). Given a population with declining household incomes, the Medi-Cal expansion to  include childless nonelderly uninsured adults may encompass a larger number of people than an

ticipated before the enactment of the ACA (Lavarreda et al. 2012). Data from CHIS 2011/2012  shows a considerable increase in the current Medi-Cal population, illustrating the result of this  income shift as the uninsured begin to take advantage of the public coverage for which they are eligible (Lavarreda and Snyder 2012). Since even a worker with wages at or near minimum wage  working full-time may be eligible for Medi-Cal under the expansion (depending on family size),  

enrollment in public health insurance programs is likely to grow even as jobs return and Califor nia climbs out of the recession. It remains to be seen if the composition of the uninsured popula tion has changed permanently or if these shifts will revert to trends seen earlier in the decade  (exemplified in the 2007 population). Still, it continues to be important for policymakers to note  the differences among county groups and to target the outreach and resources to those areas  hardest hit by difficult economic times.

8

Table 4. Odds of Being Uninsured among Nonelderly Persons, Ages 0–64, California, 2009

Variable

Odds Ratio

95 percent Confidence Limits

Recession Scale

Lower

Upper

Low Recession Impact (ref)

Moderate Recession Impact

0.868

0.705

1.068

Medium Recession Impact

1.390***

1.147

1.685

High Recession Impact

1.093

0.948

1.261

Age Group

Ages 0-18 (ref)

Ages 19-25

6.524***

5.177

8.222

Ages 26-44

4.080***

3.485

4.776

Ages 45-64

3.085***

2.605

3.652

Federal Poverty Level

401 percent FPL and Above (ref)

0%-133 percent FPL

5.825***

4.785

7.091

134%-400 percent FPL

3.434***

2.855

4.131

Work Status

Full-Time Employment (ref)

Employed, Not at Work

2.520***

1.770

3.586

Part-Time Employment

1.502**

1.210

1.865

Unemployed, Looking for Work

3.405***

2.779

4.172

Unemployed, Not Looking for Work

1.009

0.878

1.159

Citizenship and Immigration Status

US Born or Naturalized Citizen (ref)

Non-Citizen with Green Card

1.732***

1.378

2.175

Non-Citizen without Green Card

3.070***

2.539

3.714

Gender

Male (ref)

Female

0.695***

0.611

0.790

*p < .05; **p < .01; ***p < .001 

Sources: 2007 and 2009 California Health Interview Surveys, CA EDD unemployment data 9

References

Andersen, R. M. “Revisiting the Behavioral Model and Access to Medical Care: Does It Matter?” Journal of Health and Social Behavior 36 (March 1995): 1–10.  

Bohn, S., and E. Schiff. The Great Recession and Distribution of Income in California. PPIC  Publication, 2011. Retrieved from <http://www.ppic.org/main/publication.asp?i=965>. California Employment Development Department. Unemployment rate data by county [database  online]. Sacramento, CA: California Employment Development Department, 2010.  Christianson, J. B., H. T. Tu, and D. R. Samuel. Employer-Sponsored Health Insurance: Down  but Not Out. 2011. Retrieved from <http://www.hschange.com/ CONTENT/1251/>. Doty, M. M., S. D. Rustgi, C. Schoen, and S. R. Collins. Maintaining Health Insurance During a  Recession: Likely COBRA Eligibility. 2009. Retrieved from  

<http://www.commonwealthfund.org/Publications/Issue-Briefs/2009/Jan/Maintaining Health-Insurance-During-a-Recession--Likely-COBRA-Eligibility.aspx>.

Freeman, H., R. J. Blendon, L. H. Aiken, S. Sudman, C. F. Mullinix, and C. R. Corey. “Ameri cans Report on Their Access to Health Care.” Health Affairs 6, no. 1 (1987): 6–8.  doi:10.1377/hlthaff.6.1.6.

Fronstin, P. The Impact of the Recession on Employment-Based Health Coverage. 2010. Re trieved from <http://www.ebri.org/ pdf/briefspdf/EBRI_IB_05-2010_No342_Recssn-Hlth  Bens.pdf>.

Gilmer, T. P., and R. G. Kronick. “Hard Times and Health Insurance: How Many Americans  Will Be Uninsured by 2010?” Health Affairs (Project Hope) 28, no. 4 (2009): w573–7.  doi:10.1377/hlthaff. 28. 4.w573.

Gould, E. “Employer-Sponsored Health Insurance Coverage Continues to Decline in a New  Decade.” Economic Policy Institute. 2005. Retrieved from <http://www.epi.org/publication/ bp353-employer-sponsored-health-insurance-coverage/>.

Greenstein, R., S. Parrott, and A. Sherman. “Poverty and Share of Americans without Health In surance Were Higher in 2007—And Median Income for Working-Age Households Was  Lower—Than at the Bottom of Last Recession Center on Budget and Policy Priorities.” 2008. Retrieved from <http://www.cbpp.org/cms/ ?fa=view&id=621>.

Holahan, J., L. Clemans-Cope, E. Lawton, and D. Rousse. “Medicaid Spending Growth over the  Last Decade and the Great Recession, 2000–2009.” Issue Paper 8152. Kaiser Family Founda tion. 2011. Retrieved from <http:// kaiserfamilyfoundation.files.wordpress.com/2013/ 01/8152.pdf>.

Lavarreda, S.A., L. Cabezas, K. Jacobs, D. H. Roby, N. Pourat, and G. F. Kominski. The State of  Health Insurance in California: Findings from the 2009 California Health Interview Survey.  Los Angeles, CA: UCLA Center for Health Policy Research, 2012.

Lavarreda, S. A., and S. Snyder. Job-Based Coverage Insures Less Than Half of Nonelderly Cal ifornians in 2011. Los Angeles: UCLA Center for Health Policy Research, 2012. Levit, K. R., T. L. Mark, R. M. Coffey, S. Frankel, P. Santora, R. Vandivort-Warren, and K.  Malone. “Federal Spending on Behavioral Health Accelerated During Recession as Individu als Lost Employer Insurance.” Health Affairs (Project Hope) 32, no. 5 (2013): 952–62.  doi:10.1377/hlthaff.2012.1065.

Redlener, I. and R. Grant. “America’s Safety Net and Health Care Reform — What Lies Ahead?” NEJM 361 (2009): 2201–2204.

Roehr, B. US Workers Bear Cost Increase in Employer-Based Health Insurance. BMJ 341 (Sept.  10

6, 2010 1), c4846–c4846. doi:10.1136/bmj.c4846.

Ross, D. C., and C. Marks. “Challenges of Providing Health Coverage for Children and Parents  in a Recession: A 50 State Update on Eligibility Rules, Enrollment and Renewal Procedures,  and Cost-Sharing Practices in Medicaid and SCHIP in 2009.” The Henry J. Kaiser Family  Foundation. 2009. Retrieved from <http://kff.org/medicaid/report/challenges-of-providing

health-coverage-for-children/>.

Rowland, D. “Health Care and Medicaid—Weathering the Recession.” New England Journal of  Medicine 360 (2009): 1273–1276. doi:10.1056/NEJMp0901072.

Shi, L., B. Starfield, B. Kennedy, and I. Kawachi. “Income Inequality, Primary Care, and Health  Indicators.” The Journal of Family Practice 48, no. 4 (1999): 275–84. Retrieved from  <http://www.ncbi.nlm.nih.gov/pubmed/10229252>.

US Bureau of Labor Statistics. “Income, Poverty, and Health Insurance Coverage in the United  States: 2009.” Retrieved from <http://www.bls.gov/>. 2010. doi:P60-238.

Young, K., R. Garfield, L. Clemans-Cope, E. Lawton, and J. Holahan. “Enrollment-Driven Ex penditure Growth: Medicaid Spending During the Economic Downturn, FY 2007-2011.” Kaiser Family Foundation. 2013. Retrieved from <http://kff.org/medicaid/report/enrollment driven-expenditure-growth-medicaid-spending-during/>.

11

HESC 349 Final Project Spring 2016

For the Final Project, submit a 3-5 page paper presenting SPSS data runs and your interpretation of those results. Use the variables from your proposed research study  in the class midterm, from the 2014 California Health Interview Survey (CHIS) for  adolescents (ages 12-17). The SPSS dataset has been reposted in the Final Project section in the Titanium classroom.  

You can and should use the textbook and the course materials to help you prepare your  final project, but you may not work with other students. Plagiarism will be reported to  campus officials. You may ask questions about your approach to the Final Project to Dr.  Charles until Wednesday, 5/11/15, at 5 pm, either by e-mail or in office hours. The  Final Project must be submitted by Friday, 5/13/15, at midnight. Submit as a Word or  PDF document, single-spaced with a skipped space between paragraphs, with a 1 inch  margin and 12-point font.  

Your Final Project should contain the following sections (150 points total):

1) Presentation of your research question, clearly identifying the DV and the main IV  of interest. State the null and alternative hypotheses. (½ page, 20 points). 2) T-test OR chi-square test results. Choose the relevant test based on your DV and  main IV, and run it in SPSS. Cut and paste the results into your Final Project  submission. Discuss the results, including a justification for why this test was chosen,  and an identification of the p-value (1 to 1½ pages, 35 points).

3) Two regression model results. Choose either linear OR logistic regression based on  your DV, and run two regressions in SPSS. Cut and paste the relevant result tables  from both models into your Final Project submission. The first regression should be a  simple one (DV and main IV only), and the second regression should be a  multivariate one (DV, main IV, and two other IVs). Discuss the results, including a justification for why this type of regression was chosen, identification of the p-values  in each model, and a statement about what each test shows about rejecting or  failing to reject the null hypothesis. Discuss whether the bivariate model results  changed when the other two IVs were included in a multivariate model. (1 to 2  pages, 70 points).

4) State a clear, substantive conclusion based on your findings. What do the tests  tell you about the relationship between the DV and the main IV? Include a discussion  of the r-squared value and what that says about your model (½ to 1 page, 25 points).

HES­349

Final Project

May 13TH 2016

1) Presentation of your research question  

From given SPSS dataset from the 2014 California Health Interview Survey (CHIS)  for adolescents (ages 12­17), I will be using the dependent variable, X1, overweight or obese  (OVRWT2) whereas the main independent variable, X2, will be the number of sweet fruit and sport cans drunk yesterday (TC28B_P1). Also, the dependent variable is a nominal, and the  independent variable is an ordinal.

The null hypothesis (H0) indicates that two factors; overweight or obese (X1) and the  number of sweet fruit and sport cans drunk yesterday (X2); are independent each other, which also means there is no significant association between two variables. Otherwise, the null  hypothesis could be written as H0: X1 = X2.

On the other hand, the alternative hypothesis (HA) would be there is significant  relationship between overweight and obese (X1) and the number of sweet fruit and sport cans  drunk yesterday (X2); therefore, two factors are associated together, not independent. In short, the alternative hypothesis would indicate HA: X1 ≠ X2.

2) T-test OR chi-square test results  

Notice that the dependent variable (overweight or obese, X1) is a nominal whereas  the main independent variable (the number of sweet fruit and sport cans drunk yesterday X2)  is an ordinal; therefore, the cross­tab of the two variables and the chi­square test would be  used. Also, the contingency table shows the results for comparing these two variables.

OVERWEIGHT OR OBESE (CDC 2010 RECOMMENDATIONS) * # CANS OF SWEET FRUIT/SPORT DRUNK YESTERDAY (PUF 1 YR RECODE) Crosstabulation

# CANS OF SWEET FRUIT/SPORT DRUNK

YESTERDAY (PUF 1 YR RECODE)

1 CAN

186

70.7%

2 CANS

52

70.3%

3+ CANS

24

63.2%

0 CANS Total

OVERWEIGHT OR  OBESE (CDC 2010  RECOMMENDATIONS)

NO Count 485 747 % within # CANS 

OF SWEET 

FRUIT/SPORT  DRUNK 

YESTERDAY (PUF 1 YR RECODE)

71.6% 71.0%

YE Count 192 77 22 14 305

S % within # CANS  OF SWEET 

FRUIT/SPORT 

DRUNK 

YESTERDAY (PUF 1 YR RECODE)

29.3%

29.7%

36.8%

28.4% 29.0%

Total Count 677 263 74 38 1052 % within # CANS 

OF SWEET 

FRUIT/SPORT  DRUNK 

YESTERDAY (PUF 1 YR RECODE)

100.0% 100.0% 100.0% 100.0% 100.0%

Chi­Square Tests

Value df

Asymptotic Significance (2­ sided)

Pearson Chi­Square 1.299a 3 .729 Likelihood Ratio 1.248 3 .742 Linear­by­Linear Association .884 1 .347 N of Valid Cases 1052

a. 0 cells (0.0%) have expected count less than 5. The minimum expected count is 11.02.

From the results of the chi­square test, the p­value is 0.729. Since no   is given, we  α would assume   = 0.05. The p­value = 0.729,    α α ≺ p­value. Therefore, since α≺ p­value, we  would fail to reject the null hypothesis, which means the two factors; overweight or obese  (X1) and the number of sweet fruit and sport cans drunk yesterday (X2); are independent each  other. At a 5% level of significance, from the data, there is sufficient evidence to conclude  that there is no significant relationship between overweight or obese and the number of sweet fruit and sport cans drunk yesterday.

3) Two regression model results

Based on the variables, we would notice that the dependent variable (overweight or  obese, X1) is a nominal, and the main independent variable (the number of sweet fruit and  sport cans drunk yesterday, X2) is an ordinal; therefore, I will be using the bivariate (or  “simple”) logistic regression results for describing the association between overweight or  obese and the number of sweet fruit and sport cans drunk yesterday. Later on, I will also be  running the multivariate logistic regression by using two other independent variables; height  in inches, (HGHTI_P) and family size including all people supported by household  (FAMSIZE).

1. The bivariate (or “simple”) logistic regression results for describing the association between overweight or obese and the number of sweet fruit and sport cans drunk yesterday

Model Summary

Step ­2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 1265.911a.001 .001 a. Estimation terminated at iteration number 3 because parameter estimates changed by less than .001.

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

Step 1a TC28B_P1 .080 .086 .884 1 .347 1.084 Constant ­.937 .081 133.088 1 .000 .392 a. Variable(s) entered on step 1: TC28B_P1.

In the bivariate logistic regression model, the value of the odds ratio the number of  sweet fruit and sport cans drunk yesterday (TC28B_P1) is 1.084 and the p­value is 0.3947. When the alpha is set at 0.05, we would conclude that the bivariate logistic regression model  shows failing to reject the null hypothesis – overweight or obese (X1) and the number of  sweet fruit and sport cans drunk yesterday (X2); are independent each other.

2. The multivariate logistic regression results for describing the association between  overweight or obese and the number of sweet fruit and sport cans drunk yesterday,  controlling for the following additional independent variables: 1) height in inches, and 2)  family size including all people supported by household

Model Summary

Step ­2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 1265.275a.001 .002 a Estimation terminated at iteration number 4 because parameter estimates changed by less than .001.

Variables in the Equation

B S.E. Wald df Sig. Exp(B)

Step 1a TC28B_P1 .079 .086 .851 1 .356 1.082 HGHTI_P .004 .015 .083 1 .773 1.004 FAMSIZE2_P1 ­.040 .056 .508 1 .476 .961 Constant ­1.058 1.015 1.087 1 .297 .347 a. Variable(s) entered on step 1: TC28B_P1, HGHTI_P, FAMSIZE2_P1.

In the multivariate logistic regression, the p­values of the independent variables are  accordingly; 0.356 for the number of sweet fruit and sport cans drunk yesterday, 0.733 for 

height in inches, and 0.476 for family size including all people supported by household. We would notice that none of the independent variables have the p­values less than 0.05;  therefore, we could conclude that they are not significantly association with the dependent  variable; overweight or obese. 

Also, when the other two independent variables were included in a multivariate  model, the bivariate model results were accordingly changed; the p­value from 0.347 to  0.356. However, the r­squared value constantly remained the same, 0.001.

4) State a clear, substantive conclusion based on your  findings

The p­value of an independent variable (X2: the number of sweet fruit and sport cans  drunk yesterday) in a logistic regression equation shows the significance of the association  with the dependent variable (X1: overweight or obese), holding all other independent  variables constant. By given the result of the p­value; that is, 0.356, we would fail to reject  the null hypothesis, which indicates that there is no statistically significant association  overweight or obese and the number of sweet fruit and sport cans drunk yesterday. Indeed, a  larger (insignificant) p­value suggests that changes in the predictor are not associated with  changes in the response.

California State University Fullerton 

Department of Health Science 

HESC 349 - Measurement and Statistics in Health Science

General Information

Course Meeting Times: Online  

Instructor: Shana Alex Charles, MPP, PhD

Office: KHS 237

Office hours: Monday: 10 am – Noon in KHS 237 (professor’s office) Wednesday: 2:30 – 4 pm in KHS 272 (computer lab)

Phone: (657) 278-8436

Email: scharles@fullerton.edu 

Course Description  

This course is designed to provide students with an understanding of measurement theory and  statistics commonly used in the evaluation of human health and performance. An emphasis is  placed on the analysis and interpretation of data in different environments including the SPSS  statistical analysis program. Through various class activities including lecture, discussion,  evaluation, and individual projects, students will develop an understanding of statistics as  applied to research.

This course satisfies General Education Category B.5; Implications and Explorations in Mathematics  and the Natural Sciences. Courses in this subarea draw upon, integrate, apply, and extend  knowledge and skills previously acquired in subareas B1-4. These courses have a substantial  scientific and/or mathematical content and require completion of appropriate courses in subareas  B1-4 as prerequisites to enrollment. Students taking courses in subarea B5 shall:

a. Integrate themes in mathematics and/or science from cross-disciplinary perspectives.  b. Solve complex problems that require mathematical and/or scientific reasoning.  c. Relate mathematics and/or science to significant social problems or to other related  disciplines.  

d. When deemed appropriate, apply disciplinary concepts from mathematics and the natural  sciences in a variety of settings, such as community-based learning sites and activities.

Course Objectives  

To develop within the student:

1. A basic understanding of the theory and practical application of statistics in health  science research.

2. Familiarity with measures used in health science surveys (e.g., demographic questions,  measures of behavior).

3. The ability to develop your own testable research hypothesis.

4. The ability to summarize data, describe data, and test research hypotheses by  performing statistical tests using a computer program (SPSS) for a project.  

5. The ability to interpret the results of statistical findings in a meaningful way, with  acceptable and logical grammar and spelling, as indicated in the GE learning goals.  

Course Resources

1. OpenStax College, Introductory Statistics. OpenStax College. 19 September 2013.  <http://cnx.org/content/col11562/latest/>. This is the textbook for the course. It has  been posted in the classroom for free download.

2. SPSS Statistical Software

We will use this software for many assignments in the course, and you will need it for the  Final Project. You must make sure to have access to SPSS, through at least one of the  following ways:

1) Free download onto your own computer. The link is found on the portal  welcome page: sts.fullerton.edu/software. The license for using SPSS will last  through June 30, 2016.

2) You can set up timed, virtual access to SPSS on your computer by making a  request to the Pollack Library using the following website:  

http://www.fullerton.edu/it/services/vcl/. Note: If you are a Mac user, you  will need to install an additional program to your computer in order to use  virtual SPSS.  

3) You can physically go to Pollack Library, and use the computers there. All of  the computers there now have SPSS on them. You just need to go to Start ???? Programs --???? IBM SPSS Statistics.

Materials

Computer: This course is entirely online, meaning that students will need consistent access to a  computer capable of accessing online materials through Titanium. Minimum hardware and  software specifications are the following:

1. A live Internet connection. Students will need access to the Internet using a telephone  connection, DSL, or cable modem.  

2. A speaker, to listen to lectures.

3. A monitor capable of displaying information at least 800x600 pixels.  

4. If you have a Windows PC, use the Microsoft Internet Explorer Web browser (5.0 or  higher) or Netscape Navigator 7.1 only. Mac users can use the Safari Web browser.  

5. Although it is not a requirement, students should have Microsoft Office 2000 or Office  XP in addition to Adobe Acrobat Reader (download free from www.adobe.com) installed  to view and access documents provided by the instructor.

Calculator: Students will need a scientific calculator for use on homework assignments. If you do  not wish to purchase a calculator, you can search for one on Google online, and use that for  free.

Flash Drive: Or some other equivalent storage device is recommended to back up documents  and your completed assignments.  

Instructor Responsibilities  

E-mail: While I normally respond to all student inquiries within 24 hours, please expect that  sometimes, it may take me up to 24 hours to get back to you. Therefore students should plan  accordingly. If you are close to the deadline for an assignment, waiting until the day it’s due to  ask questions is not recommended as I may not be able to respond in time.

Grading: Grades will be posted during the week following their due date as much as possible. Students will be informed if there is any delay in grading.

Office Hours: Monday office hours will be held in KHS 237. Students are welcome to stop by any  time from 10 am to noon, and will be seen on a first-come, first-served basis. Wednesday office  hours will be held in KHS 272, which is an open-use computer lab. During this time, I will be  available to walk through and discuss questions about the week’s assignment.

Other Responsibilities: I will provide students with specific details regarding all assignments,  readings, and additional resources. I will actively participate in the Discussion Forums,  responding to as many as possible of the students’ initial posts each week. Additionally, I will  make myself available (either in person, on the phone, or online) to answer any questions you  may have regarding the content of the class.  

Student Technical Support

Students who have technical issues with Titanium or installing SPSS should contact the CSUF IT Help Desk at helpdesk@fullerton.edu or (657) 278-8888. They are very responsive and helpful!

Work Submission

All assignments must be submitted through Titanium, through the submission link provided. It is  always a good idea for students to keep a copy of all work submitted. Sometimes work  accidentally does not get submitted correctly. If this happens, I will ask you to e-mail me the  document with evidence that the document was last modified on the date that the assignment  was due. In the event of a technical issue with Titanium, I will arrange for alternative submission  procedures.  

Communication

Your Titanium account is linked to your CSUF e-mail. Should you want to use a different address  for communication, it is your responsibility to change it. Should you encounter any problems  with this process or any other email communication, the Help Desk should be contacted and can  be reached at helpdesk@fullerton.edu or (657) 278-8888.

Your e-mail account is your responsibility. If your account experiences problems, or is full and I  receive a returned e-mail, it is not my responsibility to resend it. Please make sure to contact me  with any problems, questions, concerns or suggestions. If several people have the same  question, I may post an announcement on the News Forum to answer this question and then all  students can learn from it. You can also post public questions for the entire class in the Help  Forum.

Please check all graded exams and assignments promptly after scores are posted on Titanium. If  you have any questions, please notify me and we can discuss the situation.

Netiquette

Just as in a classroom, students are expected to maintain an air of respectful responsibility when  participating in an online section of a class. The term netiquette refers to a set of behaviors  appropriate for online activity, including email. The core principles of netiquette can be found at  http://www.albion.com/netiquette/corerules.html. Please read these rules carefully so that you  may fully understand what behavior is expected for this course.  

Academic Dishonesty  

California State University has a strict policy against academic fraud and dishonesty. It is the  student’s responsibility to adhere to the guidelines regarding academic dishonesty as set forth  in the university handbook. A copy of the university handbook can be found at:  http://www.fullerton.edu/handbook/policy/discipline.htm 

Students who choose to participate in academic dishonesty will receive a 0 for the assignment  for the first offense, and an F in the course for the second offense. If an F is awarded for this  reason, the university administration will be contacted and the infraction will dealt with at that  level, which could include suspension or expulsion. For more information about academic  integrity policies and procedures see: www.fullerton.edu/deanofstudents/judicial/policies.htm 

Grading

The overall semester grade will be determined by calculating the total number of points earned  for Discussions, Assignments, your Midterm and Final Project. Grades will be entered into  Titanium as soon as possible and on a regular basis.  

15 Discussion Forums @ 20 points each 300 Points

13 Assignments @ 25 points each 300 Points

Midterm (Research Project Proposal) 100 Points  

Final Project (Research Project) 150 Points

Final (Timed test) 150 Points  

Total 1000 Points

Grading Criteria  

Letter grade

Percentage

Grade Point

 A+

98-100%

4.0

 A

92-97.9%

4.0

 A-

90-91.9%

3.7

 B+

88-89.9%

3.3

 B

82-87.9%

3.0

 B-

80-81.9%

2.7

 C+

78-79.9%

2.3

 C

72-77.9%

2.0

 C-

70-71.9%

1.7

 D+

68-69.9%

1.3

 D

62-67.9%

1.0

 D-

60-61.9%

0.7

 F

<60%

0.0

Discussion Forums (20 points each, 300 points total)

Note: A GPA of 2.0 or higher  is needed to satisfy certain GE  requirements and the upper division writing requirement.

Each week, there will be a Discussion forum, which will include a prompt that will require you to  respond in 300-500 words (see the Course Calendar). This “initial post” will be due on  Tuesday of that week. You are also required to review the posts of your fellow students, and to  respond to at least one of them with a substantive comment that builds on their post to move the Discussion forward. This “response post” is due on Thursday of that week. Points will be  awarded on three dimensions: 1) content of the initial post (10 points), 2) content of the  response post (5 points), and 3) academic writing style (5 points).

Assignments (25 points each, 300 points total)

Most weeks, an Assignment pertaining to that week’s readings and content will be assigned (see  the Course Calendar). This Assignment is due with an online submission by Thursday of that  week. Students may study in groups for these Assignments, but each person must submit their  own work. My Wednesday office hours will also go over the weekly Assignment in a computer  lab, allowing for in-person questions. Submission in Word is preferred, but PDFs will also be  accepted. Assignments will be graded by allotting the 25 points equally among the number of  questions. The lowest grade of the thirteen assignments will be dropped.

Midterm – Research Project Proposal (100 points)

Each student will submit an original proposal for a research study using a dataset provided by  the Instructor. The proposal will be 2-3 pages, on a health-related topic from the dataset (topic  sign-ups in Week 8). Submissions should be single-spaced, with a skipped space between  paragraphs, using 12-point font. Additional detailed information will be posted in Week 9,  including the grading rubric. The midterm will be due on Friday of Week 9.

Final Research Project (150 points total)

Based on their proposal from the midterm, students will be required to prepare and submit a  research report. This assignment requires the organization and expression of complex ideas and  the use of the SPSS skills you have learned in this class. Additional detailed information will be  posted in Week 15, including the grading rubric. The Final Project will be due on Friday of  Week 15.  

Final Timed Test (150 points total)

A timed, comprehensive Final Test will be posted in the classroom on Saturday of Finals Week.  Once begun, students will have 2 hours to take the Final. Students may use the course materials to assist them on the test, but cannot stop the Final once it is begun. The Final Test will contain  25 multiple choice questions (4 points each, for 100 points total), and 5 short answer questions  

(10 points each, for 50 points total). The Final Test can be completed at any time during  Finals Week, but must be started by 3 pm on Friday, to be completed by 5 pm.  

Make Up Policy & Late Assignment Policy  

Late work will earn a 20% point penalty for each day late, with no late work accepted after three  days. Exceptions to this rule may include a documented personal emergency, family emergency,  serious health problem etc. These will be considered on a case-by-case basis, and will be the  only reason why make-up work may be submitted. Any student in such a situation should email  me as soon as possible.  

Special Circumstances  

Students with special circumstances such as participation in collegiate athletics requiring travel  or attendance to professional conferences are requested to let the instructor know as soon as  possible. This will allow for appropriate alternate plans to be made, usually turning the work in  early prior to the conflict.  

Students with Disabilities  

According to the California State University Policy, students with disabilities need to document  their disabilities at the Disability Support Office. For more information please refer to:  http://www.fullerton.edu/dss/apply/ . Accommodations will be made as approved by the Disability  Support Office.  

Emergency

To prepare for an emergency situation, students should review: http://www.fullerton.edu/  emergencypreparedness/ep_students.html.

Class Calendar

Week

Lecture topics (Readings and  Video)

Activities (Due before midnight on date  specified unless noted otherwise)

1/23 – 1/29 Week 1

Sampling and Data

Watch: Week 1 video

Read: Chapter 1.1-1.4 (pp 9-40)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 1 (Chapter 1.6 - Sampling Experiment, pp 43-45) due on  Thursday

1/30 – 2/5

Week 2

Descriptive Statistics

Watch: Week 2 video

Read: Chapter 2.3-2.7 (pp 83-119)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 2 (Chapter 2.8 Descriptive Statistics, pp 119-120) due on  Thursday

2/6 – 2/12

Week 3

SPSS and Descriptive Statistics Watch: Week 3 video

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 3 (see template posted in classroom) due on Thursday

2/13 – 2/19 Week 4

Probability  

Watch: Week 4 video

Read: Chapter 3.1-3.5 (pp 163- 194)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 4 (Chapter 3.6 Probability Topics, pp 195-197) due on  Thursday

2/20 – 2/26 Week 5

Discrete and Continuous Random  Variables

Watch: Week 5 Video

Read: Chapter 4.1-4.3 (pp 225- 240) and Chapter 5.1-5.3 (pp 289- 313)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 5 (Homework  Chapter 4.1 and 4.2,#69-72, pp 267-268;  4.3, #82-85; pp 287; and 5.1, #72-73; pp  326) due on Thursday

2/27 – 3/4

Week 6

The Normal Distribution and the  Central Limit Theorem

Watch: Week 6 video

Read: Chapter 6.1-6.2 (pp 339- 352) and Chapter 7.1-7.3 (pp 371- 387)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 6 (Practice  Chapter 7.1, #1-6; pp 396-397) due on  Thursday

3/5 – 3/11

Week 7

Confidence Intervals

Watch: Week 7 video

Read: Chapter 8.1-8.3 (pp 411- 433)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 7 (Chapter 8.4 Confidence Interval, Home Costs, pp 434- 436) due on Thursday

3/12 – 3/18 Week 8

One Sample Hypothesis Testing (Z-test, T-test, and Probability  Functions)

Watch: Week 8 video

Read: Chapter 9.1-9.5 (pp 469- 493)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 8 (Chapter 9.6 Hypothesis Testing of a Single Sample, pp  494-496) due on Thursday

Sign up for Midterm by Friday.

3/19 – 3/25 Week 9

Midterm Week

Watch: Week 9 video

Read: Midterm Prompt

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Midterm: Due on Friday

3/26 – 4/1

Spring Recess – NO CLASS

Spring Recess – NO CLASS

4/2 - 4/8

Week 10

Two Sample Hypothesis Testing  (T-test)

Watch: Week 10 video

Read: Chapter 10.1-10.2 (pp 525- 536)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 9 (see template  posted in classroom) due on Sunday

4/9 – 4/15

Week 11

The Chi-Square Distribution

Watch: Week 11 video

Read: Chapter 11.1-11.6 (pp 577- 598)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 10 (see template  posted in classroom) due on Sunday

4/16 – 4/22 Week 12

Linear Regression and Correlation

Watch: Week 12 video

Read: Chapter 12.1-12.6 (pp 631- 655)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 11 (see template  posted in classroom) due on Sunday

4/23 – 4/29 Week 13

F-Test and ANOVA

Watch: Week 13 video

Read: Chapter 13.1-13.4 (pp 691- 707)

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 12 (see template  posted in classroom) due on Sunday

4/30 – 5/6

Week 14

Logistic Regression

Watch: Week 14 video

Read: Charles and Snyder article

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Assignment: Assignment 13 (see template  posted in classroom) due on Thursday

5/7 – 5/13

Week 15

Choosing the Right Statistical Test  and the Final Project

Watch: Week 15 video

Read: Final Project prompt

Discussion: Initial Post due on Tuesday,  Response Post due on Thursday

Final Research Project: Due by Friday

5/14 – 5/20

Finals  

Week

Finals Week

Final Test: Due by Friday at 5 pm

HESC 349 Midterm Spring 2016

Using the four variables that you have been assigned, prepare a 3-5 page proposal for a  research study using the 2014 California Health Interview Survey (CHIS) for adolescents (ages  12-17) dataset. The SPSS dataset has been posted in the Midterm section in the Titanium  classroom. This edited version of the 2014 Adolescent CHIS consists of 26 variables and 1,052  teenage survey respondents. Include the following sections in your proposal:

1) Research Question (one-half page; 20 points). Describe each of your four variables,  stating: 1) what the variable measures, and 2) what kind of variable each one is (i.e.  nominal, ordinal, interval or ratio). Briefly explain how the main independent variable  could affect the dependent variable, and how the two other independent variables  might affect that relationship. Explain what stratified sampling means, and how the CHIS  stratified sample (which is stratified by county) could affect these variables.

2) Descriptive Statistics for the Variables (1-2 pages; 40 points). Show the descriptive  statistics from SPSS for your four variables (mean, standard deviation, range [minimum  and maximum]). State the distribution (either binomial or normal) for each variable. For  each of the four variables, if the variable is a nominal or ordinal variable, present a  frequency table, but if the variable is an interval or ratio variable (called “scale” in SPSS), present a histogram. For each frequency table, include a sentence identifying the mode.  

3) Preliminary Analyses: One-Sample Hypothesis Tests (1-2 pages; 40 points). Run two  separate, one-sample hypothesis tests using SPSS, checking whether the means of your  dependent variable and your main independent variable are each statistically  significantly different from 0. Present the output for each test from SPSS, stating  whether the test shows a significant difference. Justify your statement with evidence  from the test output.

You may use the textbook and the course materials to help you prepare your proposal, but you  may not work with other students. Plagiarism will be reported to campus officials. You may ask  clarifying questions about the assignment until Wednesday at 4 pm, either by e-mail or in office  

hours. Proposals must be submitted online by Friday, 3/25/16, at midnight, in Word or PDF  format, single-spaced with a skipped space between paragraphs, with a 12-point font and 1  inch margins. Late submissions will not be accepted!

1

HES­349

Midterm

Mar 25TH 2016

TC28B_P1 (# OF SWEET FRUIT/SPORT CANS DRUNK YESTERDAY) 

1. Research Question

1) There are four distinct variables; the name of the first variable, OVRWT2,  describes overweight or obese. If its value equals to 0, it means “no” overweight or obese  whereas the value, 1, indicates “yes.” This is an example of a nominal variable because it has  two categories, but which do have an intrinsic order. The different categories of a nominal  variable can also be referred to as groups or levels of the nominal variable. TC28B_P1 is the  second variable that shows the number of sweet fruit/sport cans drunk yesterday. Its ranges  are from 0 to more than 3 cans. This variable is an example of an ordinal variable because  there is a clear ordering of the variables – 0, 1, 2 and, 3. An ordinal variable is one where the  order matters but not the difference between values. The last two variables; HGHTI_P and  FAMSIZE2 are the ratio or also known as scale variables, which has a clear definition of 0.0.  When the variables equal 0.0, there is none of that variable; for example, height and weight.  HGHTI_P indicates the height­inches, and FAMSIZE2 shows family size including all  people supported by household income.

2) In terms of the relationship between the independent and dependent variable, the  idea is that one variable is the effect of another variable or, to say it another way, that one  variable precedes and/or causes another. The independent variable does not relying on the  other variable whereas the dependent variable relies on the changes in the independent  variable. Thus, the two other independent variables could influence, directly or indirectly, the relationship between the independent and dependent variables; however, their relationship  could not explained by other factor. For example, higher in height or bigger in family size do  not cause overweight or obese. 

3) In stratified sampling, the population is partitioned into non­overlapping groups,  called strata, and then some members of each stratum are randomly selected. It also means a  layer that is made up of different layers of the population; for example, selecting samples  from different countries.

4) If the CHIS sample is stratified by the country, it could affect each of four  variables in such a way that could come with statistical differences. This is because that each  stratum is distinctly proportionate to the standard deviation of the distribution of the variable. 

Larger samples are taken in the strata with the greatest variability to generate the least  possible sampling variance.

2. Descriptive Statistics for the Variables

1) Descriptive Statistic

1. OVRWT2

Descriptive Statistics 

N Range Minimum Maximum Mean Std. Deviation

OVERWEIGHT OR OBESE 

(CDC 2010 

RECOMMENDATIONS)

1052 1 0 1 .29 .454

Valid N (listwise) 1052

2. TC28B_P1

Descriptive Statistics 

N Range Minimum Maximum Mean Std. Deviation

# CANS OF SWEET 

FRUIT/SPORT DRUNK  YESTERDAY (PUF 1 YR  RECODE)

1052 3 0 3 .50 .780

Valid N (listwise) 1052

3. HGHTI_P

Descriptive Statistics 

N Range Minimum Maximum Mean Std. Deviation

HEIGHT ­ INCHES (PUF 

RECODE) 1052 29 48 77 64.44 4.550 Valid N (listwise) 1052

4. FAMSIZE2

Descriptive Statistics 

N Range Minimum Maximum Mean Std. Deviation

FAMILY SIZE: INCL ALL 

PEOPLE SUPPORTED BY  HH INCOME (PUF 1 YR  RECODE)

1052 6 2 8 3.98 1.239

Valid N (listwise) 1052

2) Because each variable of TC28B_P1, HGHTI_P, and FAMSIZE2 has more than  two values, they are considered as normal distributions; however, there are only two values  for OVRWT2. Thus, this variable can be considered as a binomial distribution.

3) Frequency Tables and Histogram

1. OVRWT2; The mode is “No,” which indicates no overweight or obese. OVERWEIGHT OR OBESE (CDC 2010 RECOMMENDATIONS) 

Frequency Percent Valid Percent Cumulative Percent

Valid NO 747 71.0 71.0 71.0 YES 305 29.0 29.0 100.0 Total 1052 100.0 100.0

2. TC28B_P1; The mode is “0” can of sweet fruit/sport drunk yesterday. # CANS OF SWEET FRUIT/SPORT DRUNK YESTERDAY (PUF 1 YR RECODE) 

Frequency Percent Valid Percent Cumulative Percent

Valid 0 CANS 677 64.4 64.4 64.4 1 CAN 263 25.0 25.0 89.4

2 CANS 74 7.0 7.0 96.4 3+ CANS 38 3.6 3.6 100.0 Total 1052 100.0 100.0

3. HGHTI_P; The mode is “4” members of family size.

4. FAMSIZE2; The mode is “61” inches of height.

3. Preliminary Analyses: One­Sample Hypothesis Tests

1) One­sample Hypothesis Test

1. OVRWT2

One­Sample Statistics

N Mean Std. Deviation Std. Error Mean

OVERWEIGHT OR OBESE (CDC 

2010 RECOMMENDATIONS)1052 .29 .454 .014

One­Sample Test

Test Value = 0

95% Confidence Interval of

the Difference

Lower Upper

OVERWEIGHT OR 

t df Sig. (2­tailed)

Mean

Difference

OBESE (CDC 2010  RECOMMENDATIONS)

20.715 1051 .000 .290 .26 .32

2. TC28B_P1

One­Sample Statistics

N Mean Std. Deviation Std. Error Mean

# CANS OF SWEET  FRUIT/SPORT DRUNK  YESTERDAY (PUF 1 YR  RECODE)

1052 .50 .780 .024

One­Sample Test

Test Value = 0

95% Confidence Interval of

the Difference

Lower Upper

# CANS OF SWEET 

t df Sig. (2­tailed)

Mean

Difference

FRUIT/SPORT DRUNK  YESTERDAY (PUF 1 YR RECODE)

20.759 1051 .000 .499 .45 .55

2) The outputs for both OVRWT2 and TC28B_P1 present that the results do not  show a significant difference from 0.

3) For the OVRWT2, we have the evidence that Sig (p­value) is 0.0, which is less  than 0.5. Also, it suggests that the number of sample we have is 1052 on which we calculated the sample mean of 0.29. Since Sig. (p­value) is less than 0.05, we reject the null hypothesis.  Therefore, we can conclude that this sample was not drawn from the population such that the  mean of overweight or obese is equal to 0. Given the date result of TC28B_P1, we can also  find the evidence that the value of Sig. is less than 0.5, which indicates the rejection of null  hypothesis. Its evidence shows that this sample was likely not drawn from the population  such that the mean of number of sweet fruits/sport drunk yesterday is equal to 0. 

Introductory Statistics

OpenStax College

Rice University

6100 Main Street MS-380

Houston, Texas 77005

To learn more about OpenStax College, visit http://openstaxcollege.org.

Individual print copies and bulk orders can be purchased through our website.

© 2013 Rice University. Textbook content produced by OpenStax College is licensed under a Creative Commons Attribution 3.0  Unported License. Under this license, any user of this textbook or the textbook contents herein must provide proper attribution as  follows:

- If you redistribute this textbook in a digital format (including but not limited to EPUB, PDF, and HTML), then you must retain on  every page the following attribution:  

“Download for free at http://cnx.org/content/col11562/latest/.”

- If you redistribute this textbook in a print format, then you must include on every physical page the following attribution: “Download for free at http://cnx.org/content/col11562/latest/.”

- If you redistribute part of this textbook, then you must retain in every digital format page view (including but not limited to  EPUB, PDF, and HTML) and on every physical printed page the following attribution:  

“Download for free at http://cnx.org/content/col11562/latest/.”

- If you use this textbook as a bibliographic reference, then you should cite it as follows: OpenStax College, Introductory  Statistics. OpenStax College. 19 September 2013. <http://cnx.org/content/col11562/latest/>.

For questions regarding this licensing, please contact partners@openstaxcollege.org.

Trademarks

The OpenStax College name, OpenStax College logo, OpenStax College book covers, Connexions name, and Connexions logo are  not subject to the license and may not be reproduced without the prior and express written consent of Rice University.

ISBN-10 1938168208

ISBN-13 978-1-938168-20-8

Revision ST-1-000-RS

OpenStax College

OpenStax College is a non-profit organization committed to improving student access to quality learning materials. Our free textbooks  are developed and peer-reviewed by educators to ensure they are readable, accurate, and meet the scope and sequence requirements  of modern college courses. Through our partnerships with companies and foundations committed to reducing costs for students,  OpenStax College is working to improve access to higher education for all.  

OpenStax CNX

The technology platform supporting OpenStax College is OpenStax CNX (http://cnx.org), one of the world’s first and largest open education projects. OpenStax CNX provides students with free online and low-cost print editions of the OpenStax College library and  provides instructors with tools to customize the content so that they can have the perfect book for their course.  

Rice University

OpenStax College and OpenStax CNX are initiatives of Rice University. As a leading  

research university with a distinctive commitment to undergraduate education, Rice  

University aspires to path-breaking research, unsurpassed teaching, and contributions to the  

betterment of our world. It seeks to fulfill this mission by cultivating a diverse community of  

learning and discovery that produces leaders across the spectrum of human endeavor.

Foundation Support  

OpenStax College is grateful for the tremendous support of our sponsors. Without their strong engagement, the goal of free access to  high-quality textbooks would remain just a dream.

Laura and John Arnold Foundation (LJAF) actively seeks opportunities to invest in organizations and thought  leaders that have a sincere interest in implementing fundamental changes that not only yield immediate gains, but  also repair broken systems for future generations. LJAF currently focuses its strategic investments on education,  criminal justice, research integrity, and public accountability.

The William and Flora Hewlett Foundation has been making grants since 1967 to help solve social and  environmental problems at home and around the world. The Foundation concentrates its resources on activities in  education, the environment, global development and population, performing arts, and philanthropy, and makes  grants to support disadvantaged communities in the San Francisco Bay Area.

Guided by the belief that every life has equal value, the Bill & Melinda Gates Foundation works to help all people  lead healthy, productive lives. In developing countries, it focuses on improving people’s health with vaccines and  other life-saving tools and giving them the chance to lift themselves out of hunger and extreme poverty. In the  United States, it seeks to significantly improve education so that all young people have the opportunity to reach  their full potential. Based in Seattle, Washington, the foundation is led by CEO Jeff Raikes and Co-chair William  H. Gates Sr., under the direction of Bill and Melinda Gates and Warren Buffett.

The Maxfield Foundation supports projects with potential for high impact in science, education, sustainability, and  other areas of social importance.

Our mission at the Twenty Million Minds Foundation is to grow access and success by eliminating unnecessary  hurdles to affordability. We support the creation, sharing, and proliferation of more effective, more affordable  educational content by leveraging disruptive technologies, open educational resources, and new models for  collaboration between for-profit, nonprofit, and public entities.

2

This content is available for free at http://cnx.org/content/col11562/1.17

Page Expired
5off
It looks like your free minutes have expired! Lucky for you we have all the content you need, just sign up here