Quiz 2 Study Guide
Quiz 2 Study Guide PSY 302
Popular in Statistical Meth Psych
Popular in Psychlogy
This 5 page Study Guide was uploaded by Mara Anderson on Saturday April 25, 2015. The Study Guide belongs to PSY 302 at University of Oregon taught by Robbie Ross in Spring 2015. Since its upload, it has received 275 views. For similar materials see Statistical Meth Psych in Psychlogy at University of Oregon.
Reviews for Quiz 2 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: 04/25/15
Quiz 2 Study Guide PSY 302 Disclaimer This document is intended as an aide in studying for the quiz Not everything on this sheet will be on the quiz and conversely there may be questions on the quiz not covered on this sheet Quiz questions are based on content covered in lecture assigned book chapters Aplia homework and labs Basic probability odds I I number of selected outcomes Probablility total number of possible outcomes robabilit o outcome occurin Odds p y f g probability of outcome NOT occuring Odds often given as a ratio desired outcomes nondesired outcomes Odds of randomly drawing an ace from a full deck 448 The relationship of probability to the normal distribution area under curve Use zscores to transform raw scores into standardized scores We can then tell what portion of the scores are higher or lower than a certain score or what area under the curve between any two scores Zscores divide the distribution into the tail small portion and body big portion Z scores and standardized distributions how does standardizing a distribution affect the shape of the distribution if at all Standardizing a distribution does not change the shape of the distribution curve How do you find probability of finding scores greater than less than amp between 2 zs Using the unit normal table you can go from a zscore to a pvalue which will tell you the probability of nding that score in either the body or tail of that distribution If the desired probability is between two scores then the two probabilities can be added or subtracted when appropriate to nd the combined probability Unit normal tables and the 68 95 997 rule 682 of data will fall within i1 SD from the mean 3413 between mean amp 1 SD 954 of data will fall within i 2 SD from the mean 1359 between 1 and 2 SD 997 of data will fall within i 3 SD from the mean 228 between 2 and 3 SD Unit normal tables list positive zscores proportions are the same for corresponding negative zscores What is the sampling distribution of means A distribution curve made up of the means from various sample groups from a population This distribution will have less variability than a distribution of raw scores because the distribution of means have already gone through a process of being averaged together in various ways What is the central limit theorem and how does it relate to sampling distributions of means For populations with a mean of u and a standard deviation of o the sampling distribution of the mean for a sample of size n will 0 Have a mean of u 0 Approach a normal distribution as n increases 039 0 Have a standard error of 0M V How are sampling distributions of means different from or the same as the population distributions they are drawn from What is the shape of this distribution Normally distributed populations have normal sample distributions Small n s 23 sampling dist of means is similar to distribution of raw scores When n s are larger 1520 or 3 if really skewed sampling dist of means will have a normal distribution How variable is the sampling distribution of means What does sample size have to do with this What is sampling error What is the standard error How would you define it This distribution of means will have less variability than a distribution of raw scores because the distribution of means have already gone through a process of being averaged together in various ways Bigger the n the less variable the distribution sampling error a M standard error of the mean 0M Ilq Know the steps of hypothesis testing Step 0 Pose the research question Step 1 Formally state the Ho and HA about the inferred population Step 2 Set decision rule Step 3 Collect amp analyze the data Step 4 See if sample data are consistent with Ho Step 5 Answer the research question in APA style Why test a null hypothesis What IS a null hypothesis What is the alternative hypothesis The null hypothesis Ho states that there is no treatment effect in the population We test the null hypothesis because it is easier to disprove something than it is to prove something easier to show that something is likely to be false than that is probably true What is a critical region or region of significance What is alpha and what are decision rules Critical regions isare the areas in sampling distribution unlikely to be obtained if the null hypothesis is true What is a ztestzstatistic and what does it mean to say a test is signi cant Ztest statistic standardized mean comparisons sample mean compared to population means distance expressed in standard error units 0 Useful for knowing whether a sample differs signi cantly from a population 0 Tells how extreme your sample is compared to all possible samples from the population 0 A signi cant zstatistic tells us that the sample mean we have is not a part of the original population What in uences the likelihood of finding a significant difference To increase the zstatistic 0 Increase difference between M and u o Decrease standard error of mean decrease variability of scores and increase sample size When I talk about two possible populations in significance testing what do I mean Option 1 There is a single population and your sample is a part of it there is no signi cant difference between your sample and the population Option 2 There is a signi cant difference between your sample and the population meaning that the two groups are from separate populations y A PJN What are the assumptions of ztests Random sampling was used in sample selection u and a are known and the value of a is unchanged by the treatment The sampling distribution is normally shaped What types of errors can be made in signi cance testing and what are they called What in uences whether an error might be made How do you reduce different types of error Type I Errors alpha errors false positive OOOO Rejecting the null when it is actually true The treatment does not actually have a signi cantreplicable effect Occur 5 of the time Control by choosing a smaller alpha value like 001 Type II Errors beta errors false negative 0 Sample is from a different population but the mean is not extreme enough to reject the null There is a treatment effect but your sample was from the low end of that population Control by increasing power probability that the test will correctly reject a false null hypothesis power is the ability to nd an effect if one exists 0 Increase sample size 0 Lower alpha value What is an effect size How does statistical significance differ from practical significance What do I mean when I say a meaningful difference Effect size measures quantify the size of treatment effects tells us if the difference was small medium or large and helps to identify how meaningful the difference is Statistical signi cance means that mathematically there is a difference between the two groups practical signi cance means that if applied to real world situations there would be an important signi cance in the ndings of the study Although there can often be differences between values a meaningful difference is a difference that is larger than a technicality and is different enough to be important How do 1 and 2tailed tests differ Which is preferable and why One Tailed Tests allow you to reject the null with smaller effects critical z is lower but only in one tail 0 Directional test hypothesize a speci c direction of the effect 0 Easier to make type II errors Two Tailed Tests require a larger difference to reject null but allows you to observe a difference in either direction 0 Non directional test you can hypothesize without knowing which direction it will go 9 Positive onetailed test 5 Negative onetailed test gt Twotailed test gt quot note that for one tailed tests the critical region is de ned by 5 all on one end of the distribution while on two tailed tests the 5 is split up into 25 on the negative side and 25 on the positive side this is why it is easier to make type II errors during one tailed tests Q
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'