1. Review
Nine men and nine women are tested for their memory of a list of abstract nouns. The mean scores are Mmale = 15 and Mfemale = 17. The mean square based on both samples is MS = 25.
What is your best estimate of the (non-standardized) effect size, µfemale – µmale?
0.40
1.33
2.00
18.0

2. Review
Nine men and nine women are tested for their memory of a list of abstract nouns. The mean scores are Mmale = 15 and Mfemale = 17. The mean square based on both samples is MS = 25.
Calculate the standardized effect size.
0.08
0.10
0.40
0.49

3. Review
A sample of 16 subjects has a mean of 53 and a sample variance of 9. Calculate a 95% confidence interval for the population mean. The critical value for t (at 97.5% with 15 df) is 2.13.
[51.40, 54.60]
[48.21, 57.79]
[51.80, 54.20]
[50.16, 55.84]

4. Three Views of Hypothesis Testing
10/22

5. What are we doing? In science, we run lots of experiments
In some cases, there's an "effect”; in others there's not
Some manipulations have an impact; some don't
Some drugs work; some don't
Goal: Tell these situations apart, using the sample data
If there is an effect, reject the null hypothesis
If there’s no effect, retain the null hypothesis
Challenge
Sample is imperfect reflection of population, so we will make mistakes
How to minimize those mistakes?

6. Analogy: Prosecution
Think of cases where there is an effect as "guilty”
No effect: "innocent" experiments
Scientists are prosecutors
We want to prove guilt
Hope we can reject null hypothesis
Can't be overzealous
Minimize convicting innocent people
Can only reject null if we have enough evidence
How much evidence to require to convict someone?
Tradeoff
Low standard: Too many innocent people get convicted (Type I Error)
High standard: Too many guilty people get off (Type II Error)

7. Binomial Test Example: Halloween candy
Sack holds boxes of raisins and boxes of jellybeans
Each kid blindly grabs 10 boxes
Some kids cheat by peeking
Want to make cheaters forfeit candy
How many jellybeans before we call kid a cheater?

8. Binomial Test ? How many jellybeans before we call kid a cheater?
Work out probabilities for honest kids
Binomial distribution
Don’t know distribution of cheaters
Make a rule so only 5% of honest kids lose their candy
For each individual above cutoff
Don’t know for sure they cheated
Only know that if they were honest, chance would have been less than 5% that they’d have so many jellybeans
Therefore, our rule catches as many cheaters as possible while limiting Type I errors (false convictions) to 5%
Honest
Cheaters ?
Jellybeans

9. Testing the Mean ? Example: IQs at various universities
Some schools have average IQs (mean = 100)
Some schools are above average
Want to decide based on n students at each school
Need a cutoff for the sample mean
Find distribution of sample means for Average schools
Set cutoff so that only 5% of Average schools will be mistaken as Smart
100
Average
Smart ?
p(M)

10. Unknown Variance: t-test
Want to know whether population mean equals m0
H0: m = m0 H1: m ≠ m0
Don’t know population variance
Don’t know distribution of M under H0
Don’t know Type I error rate for any cutoff
Use sample variance to estimate standard error of M
Divide M – m0 by standard error to get t
We know distribution of t exactly
p(t)
Distribution of t when H0 is True m0
p(M)
Critical value
Critical Region a

11. Two-tailed Tests H0 H1 ? Often we want to catch effects on either side
Split Type I Errors into two critical regions
Each must have probability a/2 H0 H1 ?
Test statistic
Critical value
a/2

12. An Alternative View: p-values
Probability of a value equal to or more extreme than what you actually got
Measure of how consistent data are with H0
p > 
t is within tcrit
Retain null hypothesis
p < 
t is beyond tcrit
Reject null hypothesis; accept alternative hypothesis
tcrit for a = .05
p = .03 t t
t = 2.57

13. Three Views of Inferential Statistics
Confidence interval M
Effect size & confidence interval
Values of m0 that don’t lead to rejecting H0
Test statistic & critical value
Measure of consistency with H0
p-value & a
Type I error rate
Can answer hypothesis test at any level
Result predicted by H0 vs. confidence interval
Test statistic vs. critical value
p vs.  m0 m0 m0 m0
Test statistic
Critical value 1 a p

14. Review
Which of the following would lead you to reject the null hypothesis?
t < tcrit
p < a
M outside the confidence interval
t > a
p > tcrit

15. Review
In a paired-samples experiment, you get a confidence interval of [-2.4, 3.6] for µdiff. What is your conclusion?
t < tcrit, p > a, and you keep the null hypothesis
t < tcrit, p < a, and you reject the null hypothesis
t > tcrit, p > a, and you reject the null hypothesis
t > tcrit, p < a, and you keep the null hypothesis

16. Review
In a one-tailed t-test, you get a result of t = The critical value is tcrit = What region(s) correspond to alpha?
Blue
Blue + red
Blue + green
Blue + red + orange + green t
(from data)
tcrit