S 519 Evaluation of Information Systems Social Statistics

S 519: Evaluation of Information Systems Social Statistics Inferential Statistics Chapter 9: t test

This week l l What is t test Types of t test TTEST function T-test Tool. Pak

Why not z-test l l l In most cases, the z-test requires more information than we have available We do inferential statistics to learn about the unknown population but, ironically, we need to know characteristics of the population to make inferences about it Enter the t-test: “estimate what you don’t know”

William S. Gossett l l l Employed by Guinness Brewery, Dublin, Ireland, from 1899 to 1935. Developed t-test around 1905, for dealing with small samples in brewing quality control. Published in 1908 under pseudonym “Student” (“Student’s t-test”)

Types of t test

Degrees of Freedom and t test l l l Degrees of freedom describes the number of scores in a sample that are free to vary. degrees of freedom = df = n-1 The larger, the better

One sample t test l l l Very similar like z test Use sample statistics instead of population parameters (mean and standard deviation) Evaluate the result through t test table (Table B 2) instead of z test table (Table B 1)

An example l l l We show 26 babies the two pictures at the same time (one w/ his/her mother, the other a scenery picture) for 60 seconds, and measure how long they look at the facial configuration. Our null assumption is that they will not look at it for longer than half the time, μ = 30 Our alternate hypothesis is that they will look at the face stimulus longer b/c face recognition is hardwired in their brain, not learned (directional) Our sample of n = 26 babies looks at the face stimulus for M = 35 seconds, s = 16 seconds Test our hypotheses (α =. 05, one-tailed)

Step 1: Hypotheses l Sentence: l l l Null: Babies look at the face stimulus for less than or equal to half the time Alternate: Babies look at the face stimulus for more than half the time Code Symbols:

Step 2: Determine Critical Region l l l Population variance is not known, so use sample variance to estimate n = 26 babies; df = n-1 = 25 Look up values for t at the limits of the critical region from our critical values of t table Set α =. 05; one-tailed tcrit = +1. 708

Step 3: Calculate t statistic from sample l Central Limit Theorem l l l μ = 30 s. M=s/ =16/ = 3. 14

Step 4: Decision and Conclusion l l l The tobt=1. 59 does not exceed tcrit=1. 708 ∴ We must retain the null hypothesis Conclusion: Babies do not look at the face stimulus more often than chance, t(25) = +1. 59, n. s. , one-tailed. Our results do not support the hypothesis that face processing is innate.

Independent t test l A research design that uses a separate sample for each treatment condition is called an independent-measures (or betweensubjects) research design.

t Statistic for Independent. Measures Design l l l The goal of an independent-measures research study: To evaluate the difference of the means between two populations. Mean of first population: μ 1 Mean of second population: μ 2 Difference between the means: μ 1 - μ 2

t Statistic for Independent. Measures Design l Null hypothesis: “no change = no effect = no difference” l l H 0: μ 1 - μ 2 = 0 Alternative hypothesis: “there is a difference” l H 1: μ 1 - μ 2 ≠ 0

T test formula : is the mean for Group 1 : is the mean for Group 2 : is the number of participants in Group 1 : is the number of participants in Group 2 : is the variance for Group 1 : is the variance for Group 2 Value for degrees of freedom: df = df 1 + df 2

An Example Group 1 7 5 3 4 3 6 2 10 3 10 8 5 8 1 5 1 8 4 5 3 5 7 1 9 2 5 2 12 15 4 Group 2 5 4 4 5 5 7 8 8 9 8 3 2 5 4 4 6 7 7 5 6 4 3 2 7 6 2 8 9 7 6

T test steps l Step 1: A statement of the null and research hypotheses. l Null hypothesis: there is no difference between two groups l Research hypothesis: there is a difference between the two groups

T test steps l Step 2: setting the level of risk (or the level of significance or Type I error) associated with the null hypothesis l 0. 05

T test steps l Step 3: Selection of the appropriate test statistic l l Determine which test statistic is good for your research Independent t test

T test steps l Step 4: computation of the test statistic value l t= 0. 14

T test steps l Step 5: determination of the value needed for the rejection of the null hypothesis l l Table B 2 in Appendix B (S-p 360) Degrees of freedom (df): approximates the sample size l l l Group 1 sample size -1 + group 2 sample size -1 Our test df= 58 Two-tailed or one-tailed l l Directed research hypothesis one-tailed Non-directed research hypothesis two-tailed

T test steps l Step 6: A comparison of the obtained value and the critical value l l l 0. 14 and 2. 001 If the obtained value > the critical value, reject the null hypothesis If the obtained value < the critical value, retain the null hypothesis

T test steps l Step 7 and 8: make a decision l What is your decision and why?

Interpretation l How to interpret t(58) = 0. 14, p>0. 05, n. s.

Excel: TTEST function l TTEST (array 1, array 2, tails, type) l l array 1 = the cell address for the first set of data array 2 = the cell address for the second set of data tails: 1 = one-tailed, 2 = two-tailed type: 1 = a paired t test; 2 = a two-sample test (independent with equal variances); 3 = a twosample test with unequal variances

Excel: TTEST function l l It does not compute the t value It returns the likelihood that the resulting t value is due to chance (the possibility of the difference of two groups is due to chance)

Excel Tool. Pak l Select t-test: two sample assuming equal variances t-Test: Two-Sample Assuming Equal Variances Mean Variance Observations Pooled Variance Hypothesized Mean Difference df t Stat P(T<=t) one-tail t Critical one-tail P(T<=t) two-tail t Critical two-tail Variable 1 5. 43333 11. 70229885 30 7. 979885057 0 58 -0. 137103112 0. 44571206 1. 671552763 0. 891424121 2. 001717468 Variable 2 5. 53333 4. 257471264 30

Effect size l If two groups are different, how to measure the difference among them l Effect size ES: effect size : the mean for Group 1 : the mean for Group 2 SD: the standard deviation from either group

Effect size l A small effect size ranges from 0. 0 ~ 0. 2 l l A medium effect size ranges from 0. 2 ~ 0. 5 l l l The two groups are different A large effect size is any value above 0. 50 l l Both groups tend to be very similar and overlap a lot The two groups are quite different ES=0 the two groups have no difference and overlap entirely ES=1 the two groups overlap about 45%