Statistics for the Social Sciences Psychology 340 Fall 2006 Hypothesis testing Outline (for week) Statistics for the Social Sciences • Review of: – Basic probability – Normal distribution – Hypothesis testing framework • Stating hypotheses • General test statistic and test statistic distributions • When to reject or fail to reject Hypothesis testing Statistics for the Social Sciences • Example: Testing the effectiveness of a new memory treatment for patients with memory problems – Our pharmaceutical company develops a new drug treatment that is designed to help patients with impaired memories. – Before we market the drug we want to see if it works. – The drug is designed to work on all memory patients, but we can’t test them all (the population). – So we decide to use a sample and conduct the following experiment. – Based on the results from the sample we will make conclusions about the population. Hypothesis testing Statistics for the Social Sciences • Example: Testing the effectiveness of a new memory treatment for patients with memory problems Memory patients Memory treatment Memory 55 Test errors No Memory treatment Memory 60 errors Test • Is the 5 error difference: – A “real” difference due to the effect of the treatment – Or is it just sampling error? 5 error diff Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing – Procedure for deciding whether the outcome of a study (results for a sample) support a particular theory (which is thought to apply to a population) – Core logic of hypothesis testing • Considers the probability that the result of a study could have come about if the experimental procedure had no effect • If this probability is low, scenario of no effect is rejected and the theory behind the experimental procedure is supported Basics of Probability Statistics for the Social Sciences Possible successful outcomes Probability All possible outcomes • Probability – Expected relative frequency of a particular outcome • Outcome – The result of an experiment Flipping a coin example Statistics for the Social Sciences What are the odds of getting a “heads”? n = 1 flip Possible successful outcomes Probability All possible outcomes One outcome classified as heads Total of two outcomes = 1 2 = 0.5 Flipping a coin example Statistics for the Social Sciences n=2 Number of heads 2 1 1 What are the odds of getting two “heads”? One 2 “heads” outcome Four total outcomes = 0.25 0 This situation is known as the binomial # of outcomes = 2n Flipping a coin example Statistics for the Social Sciences n=2 Number of heads 2 1 1 0 What are the odds of getting “at least one heads”? Three “at least one heads” outcome Four total outcomes = 0.75 Flipping a coin example Statistics for the Social Sciences n=3 3= n = 2 2 8 total outcomes HHH Number of heads 3 HHT 2 HTH 2 HTT 1 THH 2 THT 1 TTH 1 TTT 0 Flipping a coin example Statistics for the Social Sciences Number of heads 3 Distribution of possible outcomes probability (n = 3 flips) .4 .3 .2 .1 .125 .375 .375 .125 0 1 2 3 Number of heads 2 X f p 3 1 .125 2 2 1 3 3 .375 .375 1 0 1 .125 1 2 1 0 Flipping a coin example Statistics for the Social Sciences Distribution of possible outcomes probability (n = 3 flips) .4 .3 .2 .1 .125 .375 .375 .125 0 1 2 3 Number of heads Can make predictions about likelihood of outcomes based on this distribution. What’s the probability of flipping three heads in a row? p = 0.125 Flipping a coin example Statistics for the Social Sciences Distribution of possible outcomes probability (n = 3 flips) .4 .3 .2 .1 .125 .375 .375 .125 0 1 2 3 Number of heads Can make predictions about likelihood of outcomes based on this distribution. What’s the probability of flipping at least two heads in three tosses? p = 0.375 + 0.125 = 0.50 Flipping a coin example Statistics for the Social Sciences Distribution of possible outcomes probability (n = 3 flips) .4 .3 .2 .1 .125 .375 .375 .125 0 1 2 3 Number of heads Can make predictions about likelihood of outcomes based on this distribution. What’s the probability of flipping all heads or all tails in three tosses? p = 0.125 + 0.125 = 0.25 Hypothesis testing Statistics for the Social Sciences Distribution of possible outcomes (of a particular sample size, n) Can make predictions about likelihood of outcomes based on this distribution. • In hypothesis testing, we compare our observed samples with the distribution of possible samples (transformed into standardized distributions) • This distribution of possible outcomes is often Normally Distributed The Normal Distribution Statistics for the Social Sciences • The distribution of days before and after due date (bin width = 4 days). -14 0 14 Days before and after due date The Normal Distribution Statistics for the Social Sciences • Normal distribution The Normal Distribution Statistics for the Social Sciences • Normal distribution is a commonly found distribution that is symmetrical and unimodal. – Not all unimodal, symmetrical curves are Normal, so be careful with your descriptions • It is defined by the following equation: 1 2 2 -2 -1 0 1 2 e (X ) 2 / 2 2 The Unit Normal Table Statistics for the Social Sciences z .00 .01 -3.4 -3.3 : : 0 : : 1.0 : : 3.3 3.4 0.0003 0.0005 : : 0.5000 : : 0.8413 : : 0.9995 0.9997 0.0003 0.0005 : : 0.5040 : : 0.8438 : : 0.9995 0.9997 • The normal distribution is often transformed into z-scores. • Gives the precise proportion of scores (in z-scores) between the mean (Z score of 0) and any other Z score in a Normal distribution – Contains the proportions in the tail to the left of corresponding z-scores of a Normal distribution • This means that the table lists only positive Z scores Using the Unit Normal Table Statistics for the Social Sciences z .00 .01 -3.4 -3.3 : : 0 : : 1.0 : : 3.3 3.4 0.0003 0.0005 : : 0.5000 : : 0.8413 : : 0.9995 0.9997 0.0003 0.0005 : : 0.5040 : : 0.8438 : : 0.9995 0.9997 50%-34%-14% rule Similar to the 68%-95%-99% rule 34.13% 13.59% -2 -1 0 1 2.28% 2 At z = +1: 15.87% (13.59% and 2.28%) of the scores are to the right of the score 100%-15.87% = 84.13% to the left Using the Unit Normal Table Statistics for the Social Sciences z .00 .01 -3.4 -3.3 : : 0 : : 1.0 : : 3.3 3.4 0.0003 0.0005 : : 0.5000 : : 0.8413 : : 0.9995 0.9997 0.0003 0.0005 : : 0.5040 : : 0.8438 : : 0.9995 0.9997 • Steps for figuring the percentage above of below a particular raw or Z score: 1. Convert raw score to Z score (if necessary) 2. Draw normal curve, where the Z score falls on it, shade in the area for which you are finding the percentage 3. Make rough estimate of shaded area’s percentage (using 50%-34%-14% rule) Using the Unit Normal Table Statistics for the Social Sciences z .00 .01 -3.4 -3.3 : : 0 : : 1.0 : : 3.3 3.4 0.0003 0.0005 : : 0.5000 : : 0.8413 : : 0.9995 0.9997 0.0003 0.0005 : : 0.5040 : : 0.8438 : : 0.9995 0.9997 • Steps for figuring the percentage above of below a particular raw or Z score: 4. Find exact percentage using unit normal table 5. If needed, add or subtract 50% from this percentage 6. Check the exact percentage is within the range of the estimate from Step 3 SAT Example problems Statistics for the Social Sciences • The population parameters for the SAT are: = 500, = 100, and it is Normally distributed Suppose that you got a 630 on the SAT. What percent of the people who take the SAT get your score or worse? z X 630 500 From the table: 1.3 100 z(1.3) =.0968 So 90.32% got your score or worse -2 -1 That’s 9.68% above this score 1 2 The Normal Distribution Statistics for the Social Sciences • You can go in the other direction too – Steps for figuring Z scores and raw scores from percentages: 1. Draw normal curve, shade in approximate area for the percentage (using the 50%-34%-14% rule) 2. Make rough estimate of the Z score where the shaded area starts 3. Find the exact Z score using the unit normal table 4. Check that your Z score is similar to the rough estimate from Step 2 5. If you want to find a raw score, change it from the Z score Inferential statistics Statistics for the Social Sciences • Hypothesis testing – Core logic of hypothesis testing • Considers the probability that the result of a study could have come about if the experimental procedure had no effect • If this probability is low, scenario of no effect is rejected and the theory behind the experimental procedure is supported – A five step program • • • • • Step 1: State your hypotheses Step 2: Set your decision criteria Step 3: Collect your data Step 4: Compute your test statistics Step 5: Make a decision about your null hypothesis Hypothesis testing Statistics for the Social Sciences • Hypothesis testing: a five step program – Step 1: State your hypotheses: as a research hypothesis and a null hypothesis about the populations • Null hypothesis (H0) This is the one that you test • There are no differences between conditions (no effect of treatment) • Research hypothesis (HA) • Generally, not all groups are equal – You aren’t out to prove the alternative hypothesis • If you reject the null hypothesis, then you’re left with support for the alternative(s) (NOT proof!) Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – Step 1: State your hypotheses In our memory example experiment: One -tailed – Our theory is that the treatment should improve memory (fewer errors). H0: Treatment > No Treatment HA:Treatment < No Treatment Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – Step 1: State your hypotheses In our memory example experiment: direction One -tailed specified – Our theory is that the treatment should improve memory (fewer errors). no direction specified Two -tailed – Our theory is that the treatment has an effect on memory. H0: Treatment > No Treatment H0: Treatment = No Treatment HA:Treatment < No Treatment HA:Treatment ≠ No Treatment One-Tailed and Two-Tailed Hypothesis Tests Statistics for the Social Sciences • Directional hypotheses – One-tailed test • Nondirectional hypotheses – Two-tailed test Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – Step 1: State your hypotheses – Step 2: Set your decision criteria • Your alpha () level will be your guide for when to reject or fail to reject the null hypothesis. – Based on the probability of making making an certain type of error Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – Step 1: State your hypotheses – Step 2: Set your decision criteria – Step 3: Collect your data Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – – – – Step 1: State your hypotheses Step 2: Set your decision criteria Step 3: Collect your data Step 4: Compute your test statistics • Descriptive statistics (means, standard deviations, etc.) • Inferential statistics (z-test, t-tests, ANOVAs, etc.) Testing Hypotheses Statistics for the Social Sciences • Hypothesis testing: a five step program – – – – – Step 1: State your hypotheses Step 2: Set your decision criteria Step 3: Collect your data Step 4: Compute your test statistics Step 5: Make a decision about your null hypothesis • Based on the outcomes of the statistical tests researchers will either: – Reject the null hypothesis – Fail to reject the null hypothesis • This could be correct conclusion or the incorrect conclusion Error types Statistics for the Social Sciences • Type I error (): concluding that there is a difference between groups (“an effect”) when there really isn’t. – Sometimes called “significance level” or “alpha level” – We try to minimize this (keep it low) • Type II error (): concluding that there isn’t an effect, when there really is. – Related to the Statistical Power of a test (1-) Error types Statistics for the Social Sciences There really isn’t an effect Reject H0 Experimenter’s conclusions Fail to Reject H0 Real world (‘truth’) H0 is correct H0 is wrong There really is an effect Error types Statistics for the Social Sciences Real world (‘truth’) I conclude that there is an effect H0 is correct Reject H0 Experimenter’s conclusions Fail to Reject H0 I can’t detect an effect H0 is wrong Error types Statistics for the Social Sciences Real world (‘truth’) H0 is correct Reject H0 Experimenter’s conclusions Fail to Reject H0 H0 is wrong Type I error Type II error Performing your statistical test Statistics for the Social Sciences • What are we doing when we test the hypotheses? Real world (‘truth’) H0: is true (no treatment effect) H0: is false (is a treatment effect) One population Two populations XA the memory treatment sample are the same as those in the population of memory patients. XA they aren’t the same as those in the population of memory patients Performing your statistical test Statistics for the Social Sciences • What are we doing when we test the hypotheses? – Computing a test statistic: Generic test Could be difference between a sample and a population, or between different samples observed difference test statistic difference expected by chance Based on standard error or an estimate of the standard error “Generic” statistical test Statistics for the Social Sciences • The generic test statistic distribution (think of this as the distribution of sample means) – To reject the H0, you want a computed test statistics that is large – What’s large enough? • The alpha level gives us the decision criterion Distribution of the test statistic -level determines where these boundaries go “Generic” statistical test Statistics for the Social Sciences • The generic test statistic distribution (think of this as the distribution of sample means) – To reject the H0, you want a computed test statistics that is large – What’s large enough? • The alpha level gives us the decision criterion Distribution of the test statistic If test statistic is here Reject H0 If test statistic is here Fail to reject H0 “Generic” statistical test Statistics for the Social Sciences • The alpha level gives us the decision criterion Two -tailed One -tailed = 0.05 Reject H0 Reject H0 0.025 split up into the two tails 0.025 Fail to reject H0 Reject H0 Fail to reject H0 Fail to reject H0 “Generic” statistical test Statistics for the Social Sciences • The alpha level gives us the decision criterion Two -tailed One -tailed = 0.05 all of it in one tail Reject H0 Reject H0 0.05 Fail to reject H0 Reject H0 Fail to reject H0 Fail to reject H0 “Generic” statistical test Statistics for the Social Sciences • The alpha level gives us the decision criterion Two -tailed One -tailed = 0.05 Reject H0 all of it in one tail Reject H0 0.05 Fail to reject H0 Reject H0 Fail to reject H0 Fail to reject H0 “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. • After the treatment they have an average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have of memory errors that is a distribution Normal, = 60, = 8? • Step 1: State your hypotheses H0: the memory treatment sample are the same as those in the population of memory patients. Treatment > pop > 60 HA: they aren’t the same as those in the population of memory patients Treatment < pop < 60 “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test H0: Treatment > pop > 60 HA: Treatment < pop < 60 Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. • After the treatment they have an average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have of memory errors that is a distribution Normal, = 60, = 8? • Step 2: Set your decision criteria = 0.05 One -tailed “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test H0: Treatment > pop > 60 HA: Treatment < pop < 60 Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. • After the treatment they have an average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have of memory errors that is a distribution Normal, = 60, = 8? One -tailed • = 0.05 Step 3: Collect your data “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test H0: Treatment > pop > 60 HA: Treatment < pop < 60 Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. • After the treatment they have an average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have of memory errors that is a distribution Normal, = 60, = 8? = 0.05 One -tailed • Step 4: Compute your test statistics zX X X X = -2.5 55 60 8 16 “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. H0: Treatment > pop > 60 HA: Treatment < pop < 60 = 0.05 One -tailed zX 2.5 • Step 5: Make a decision • After the treatment they have an about your null hypothesis average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have 5% of memory errors that is a distribution Normal, = 60, = 8? -2 -1 Reject H0 1 2 “Generic” statistical test Statistics for the Social Sciences An example: One sample z-test H0: Treatment > pop > 60 HA: Treatment < pop < 60 Memory example experiment: • We give a n = 16 memory patients a memory improvement treatment. • After the treatment they have an average score of X = 55 memory errors. • How do they compare to the general population of memory patients who have of memory errors that is a distribution Normal, = 60, = 8? One -tailed = 0.05 zX 2.5 • Step 5: Make a decision about your null hypothesis - Reject H0 - Support for our HA, the evidence suggests that the treatment decreases the number of memory errors