BNAD 276 exam 1 study guide
BNAD 276 exam 1 study guide BNAD 276 001
Popular in Statistic Inference in Management
Popular in Department
verified elite notetaker
One Day of Notes
verified elite notetaker
verified elite notetaker
verified elite notetaker
verified elite notetaker
verified elite notetaker
This 9 page Study Guide was uploaded by madelinef on Saturday October 1, 2016. The Study Guide belongs to BNAD 276 001 at University of Arizona taught by Dr. Suzanne Delaney in Fall 2016. Since its upload, it has received 10 views.
Reviews for BNAD 276 exam 1 study guide
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 10/01/16
BNAD 276 Exam 1 Study Guide Operational Definition- definition of a construct/characteristic in terms of how it is measures specifically for a particular thing o Construct- represents relatively abstract constructs o Measurements- asses’ observable characteristics or behaviors in a reduction of uncertainty Validity- are you measuring what you intend to? Reliability- the extent to which a test yields consistent results Dependent variable- the variable being measured by investigator o The Data that is being recorded Independent variable- the factor that is being manipulated (or compared) by the investigator Within participant design- each subject participated in every level of variable (repeated measures) Between participant design- each subject participated in only 1 level of independent variables. Random assignment- any subject has an equal chance of getting assigned to either condition. o If you have random assignment, you may have a true experiment and thus cause and effect o If no random assignment you have a quasi-experiment and can NOT talk about cause and effect. (gender and spending) Correlation does not affect causation Random sampling- each person in population has an equal chance of being selected to be in the sample. Nothing to do with quasi and true! Population- entire group to study Sample- subset of population who participated Cross sectional data- data collected by recording a characteristic of many subjects at the same point in time, or without regard to differences in time. Time series data- data collected by recording a characteristic of subject over several time periods Variable- when a characteristic or difference differs in kind or degree o Qualitative variable- using labels or names to identify a characteristic Discrete- assumes countable number of variables. Variables that can only assume whole numbers. Ex: eggs in a carton, number of kids Continuous- uncountable values in an interval. Variables that can assume any variable. There are an infinite number of values between any 2 numbers. Example: height and distance. Direct variable- what is being measured, data being gathered Indirect variable- what is being manipulated. Data being gathered Independent variable- what is being manipulated, how groups are different Random assignment- everyone has the exact same chance to either condition Random sampling- each person has an equal chance to either condition Ture experiment- random assignment, cause and effect Quasi experiment- no random sampling, no cause and effect Placebo single blind- you do not know what group you’re in Double blind- experimenter is blind and researches you Continuous variable- variables that can assume any value. There are an infinite number of values between any 2 numbers. Ex: height and distance Discrete variables- variables that can only assume while numbers. NO intermediate values between the whole number. Ex: number of eggs in a carton, number of children Categorical data- qualitative, a set of observable observations where any single observation is a word or number represents a class or category. Ex: gender, type of dog Numeric data- quantitative data, a set of observations where any single observation Is a number represented by a count or amount. Ex: temperature 4 levels of measurement o nominal- classification, differs in kind, names of categories. Not numeric at all! Ex: categories, names, jersey number, ethnic group, a label basically o ordinal- order, rankings, or differing in degree. More like a number but still not one. Ex: place in a race, grade in school. o Interval- measureable, differs in amount, equal intervals. Ex: sat score, GPA, shoe size o Ratio- measurable, different in amount with a true zero. Ex: time taken to do something, distance, weight, kelvin scale, bank account, number of kids Time series design- each observation represents a measurement at some point in time, repeated measurement allows us to see trends Cross sectional design- each observational represents a measurement at some point in time, comparing lets us see differences. Census- measures each person in the specific population Sample- measures a subset of population Parameter- measurement or characteristic of the population o Usually unknown (estimated) o Represented by Greek letters Statistic- numerical value calculated form a sample usually represented by roman letters o Descriptive statistic- organizing and summarizing data o Inferential statistic- generalizing beyond actual observations. Making inferences based on data collected Simple random sampling- population has equal probability of being included Sample frame - how you define population Systematic random sampling- a probability sampling technique that involves selecting every person you are sampling from Stratified sampling- sampling technique that involved dividing a sample into subgroups Cluster sampling- divide a population subgroup into subgroups. Can either measure everyone or select subjects. Non random sampling o Convenience sampling – sampling technique that involves people. A non-random sample and vulnerable to bias. Snowball sample- non-random technique in which 1 or more members of a population are located and used to lead research to another member of the population. Used when we have no way to find them Judgement sample- technique that involves sampling people who an expert said would be useful to sample Scatterplot- displays relationship between two continuous variables Correlation- measure of how two variable co-occur and also can be used for prediction o Range between -1 and +1. Closer to 0 the weaker the relationship and worse the prediction. o -1 and +1 are the same! The negative refers to if the values of 1 variable goes up the value for the other goes down. And positive values are if the one goes up, the three goes up o 0.9-0.7= strong predictor o 0.4-0.6= medium predictor o 0.1-0.3= weak predictor o 0= no relationship Correlation does not imply causation. o Is it possible that they are causally related? Yes, but the analysis does not answer that question o Even if it is a perfect correlation it still doesn’t cause causation o Ex: birthday cake doesn’t make you older Linear Vs. Curvilinear Relationships o Linear relationship- a relationship that is described with a straight line o Curvilinear relationship- a relationship that can be described best with a curved line. Not appropriate for correlation o When talking about correlation you need to say… Positive or negative (direction) Established values (actual number) Variables Strength (weak, moderate, strong) Ex: This shows a strong negative relationship (r=-0.9) between a and b. Frequency distribution- an organized list of observations and their frequency of occurrence o Guidelines for creation: o Classes should be mutually exclusive: each observation should be represented only once. No overlap between classes. Ex: political affiliation, gender, number of kids in family. o Set of classes should be exhaustive: Should include all data value possibilities! No data falls out of range. Ex: 0-3, 6-8, 9-11 is WRONG! But 0-3, 4-7, 8-11, 12-15= Right! o All classes share equal sized intervals (even if frequency for class is 0). Wrong: 0-9, 10-12, 13-19 Right- 0-4, 5-9, 10-14 o Selecting number of classes is subjective Generally, 5-15 will often work On a frequency histogram, frequency is on the y axis Pareto chart- categories are displayed in descending order of frequency Stacked bar chart- bar height is the sum of several subtotals Simple line chart- often used for time series data (continuous data) o Space between dots implies a continuous flow Pie chart- general idea of data that must sum to a total. o Use with caution! Often overused. o 2D is better than 3D usually because the third dimension doesn’t add much Frequency Distributions The normal curve- bell curve shaped Central tendency- where’s data values concentrated? What are the typical or middle data values? o Mean- the balance point of distribution. Found by adding up all observations and then dividing by the number of observations. For a sample x/n= mean For a population X/N= mean = add up, x or X= scores, n or N= number of scores o Median- middle value when observations are ordered from least to most (or most to least). Numbers must be in numerical order to start If there are 2 medians, take the average of them 1 quartile- middle number of the lower half second quartile- median 3 quartile- middle number of upper half o mode- the value of the most common observation bimodal distribution- if there are 2 most frequent observations o trimmed mean- trim off a bit at the top and bottom to get rid of outliers. Dispersion- how much variation is there in data? How spread out are the data values? How unusual are they? Standard deviation, variance, range, absolute deviation Shape- are the data values distributed symmetrically? Skewed? Sharply peaked? Bimodal? Central Tendencies for Qualitative Data Mode Is good for nominal or ordinal data Median can be used for ordinal data Mean can be used with interval or ratio data Measure of central tendency- describes how scores tend to cluster to the center o In all distributions Mode=tallest part Median=middle score Mean= balanced part Positively skewed distribution: mode<median<mean Negatively skewed distribution: mean<median<mode Means most affected by outliers or skewed distribution! Variability- some distributions are more valuable than others. o The larger the variability, the wider the curve and vice versa o Range- difference between largest and smallest number Z Scores- tell us the relative location o 100% or 1.0 under curve o 50% or 0.5 is the mean…above and below. o Why do we care about raw scores? They dictate the x-axis. Gives us the actual number! o For every raw score there’s only 1 z score Example: mean=100, standard deviation=5 so 1 standard deviation would be at 105 and 95. If score is within 2 standard deviations (z<2)= not unusual. o Middle 95% o Typical If z score is more than 2 and less than 3 (2<z<3) it is an unusual score Outlier= >3 standard deviation (ex: 3.2) Extreme outlier= >4 standard deviations (ex: 4.3) Proportions percentiles, percent of The raw scores and proportions do not correlate but they do when they are converted to Z Scores! Normal Distribution o Raw scoreZ Scoreprobabilities o Raw score=mean+ (z score)(standard deviation) o Bigger Z’s are getting farther from the mean! Practice Problem o Mean-30, standard deviation-2 finds the raw score of the 75 th percentile. 75-50=25. 30+ (0.67)(2)= 31.34 o find the 25 percentile 50-25=25 30+(-0.67)(2)= 28.66 Deviation Score- average-number. o 0 means average o a negative number can be good or bad. It depends on the situation it is in. o Example: the average is 6’0”. 6’2”= +2 above average 5’8”= -4 below average o (x-) x= your data and =average average deviation score- add up deviation scores and divide by amount. It will always be 0 unless you did it wrong. To get a standard deviation you want to square the deviation, add it up x−¿ ¿ ¿2 and square root it. ( ¿ ) ¿ ¿ Σ √¿ Standard deviation- the average amount by which observations deviate on either side of their mean 4 equations to know o Notes: o = sigma o S and = standard deviation sample (s) and population () o S^2 and ^2= variance sample (s) and population () x−¿ ¿ 2 ¿2 x−x ) o Sum of squares= ¿ and Σ n−1 Σ¿ √ ¿ o n-1= degree of separation standard deviation squared= variance n-1 is degrees of freedom o you lose a degree of freedom every time you guess (for every parameter you estimate) o it’s a mathematical correction because of estimating difference is standard deviation squared is variance. Variance square rooted is standard deviation. o Numerator= sum of squares. 1 Standard deviation above and below mean is 68% of population 2 standard deviations above and below mean is 95% of population 3 standard deviations above and below mean is 99.7% of population Z-Score o 100% under curve o 50% (0.5) Is above mean and 50% of below mean o Why do we care about raw scores? Because they give us the ACTUAL number o Proportions= percent of/ percentile o The raw scores and proportions do not correlate naturally. But they do both correlate to Z scores so we change them to that to compare! Z= x-/ Z score o 100% under curve (0.1) o 50% on half of it (0.5) Probability Empirical probability- relative frequency approach o Number of observed outcomes/number of observations o Ex: Eller admissions, rotten apples Classical probability- a priori probability based on logic rather than on data or experience. All options are equally likely o Deductive rather than inductive o Ex; coin toss, dice roll Subjective probability- based on someone’s personal judgement (usually an expert) and used when empirical or classic are not available o Ex; there is a 5% change a and b will merge. I am 90% sure I can jump over than puddle Compliment is the probability of not A. o P(A)A’=P(A’) o Ex: if there is a 5% check you get A there is 95% chance you do not get A Mutually exclusive- if the occurrence of any one of them implies the non-occurrence of the remaining characteristic o Cannot occur at the same time. Cant logically be true. o Ex: can’t be male and female, can’t have an A and a B in the same class as a final grade Collectively exhausted- if the union is the entire sample space Union- probability of A or B. P (A U B) Intersection- probability of A and B. P (A n B) o Issue: things in both groups get counted twice Conditional probability- probability that A has occurred given that B has occurred. PAlB)= P(A n B)/P(B) o Ex: population 16-21 not in college. Unemployed= 13.5%-->0.1350 Did not graduate90//05%= 0.2905 If 2 events are mutually exclusive their intersecting is a null set and we can use a special law of addition Why do we conduct research? o 1. To explore potential phenomena- explore if a phenomenon is present with a fresh take. Generate new ideas and discover relationships o 2. To describe a phenomena- build a vocabulary of constructs and make distinctions, cluster characteristics into related characteristics o 3. To explain and model phenomena- explanation to find cause and effect relationships, propose mechanisms that determine outcomes, show how o 4. To predict future behavior- what characteristics likely to result in x, explanations help with predictions but being able to predict does not equal being able to explain o 5. To influence behavior- how can we use what we know about human behavior to affect how people around us react and behave and do what we want them to do evaluating investments- the standard deviation refers to the volatility or risk of the investment. o More risklarger standard deviation o The numerator is the difference our investment is from a lowest risk type of investment Sharpe ratio- also called reward to variability ratio. Formula is o Mean-risk free mean/ standard deviation. o Parallels the same formula as z score!
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'