Part 7: Determining proper statistical test to use (10 pts total) Directions: Re
ID: 3319838 • Letter: P
Question
Part 7: Determining proper statistical test to use (10 pts total) Directions: Read each of the scenarios below and determine which of the following statistical tests that we have learned is most appropriate to test the research question. Consider things like continuous vs. nominal variables, scores vs. frequencies, number of groups, and what we know (or don't know) about various populations. Briefly explain why you chose the test that you did for each. HINT* each test in the word bank below will only be used once! Word Bank single sample Z test independent sample t test Correlation/Regression One-way ANOVA dependent sample test Two-way/Factorial ANOVA single sample t test Chi Square Test (a) A school psychologist surveyed 75 students (both boys and girls) who had been reported as "bullies" and asked them which bullying method they used most often: physical, verbal, psychological, or cyber-bullying. A school psychologist wants to know if there is statistical evidence to indicate a dependence between gender and bulying method (that is, does one gender tend to adopt one kind of bullying method over another)? Statistical test: Why: (b) An organizational psychologist wants to know if workers' health would improve if a company gave workers extra days off. All workers take a standard physical exam twice a year and are given an overall health rating. The psychologist selected five workers at random, arranged to have them receive an additional 2 days off every month over the period between health exams, and then compared the health scores on the two exams, before and after the extra days off. Statistical test: WhyExplanation / Answer
(a) Chi square test
(i)Because the variables are categorical (Male/Female) (Methods of bullying)
(ii) We are finding a relationship between gender and type of bullying.
(b) Dependent sample t test
(i) The same 5 people are being put to the test before and after 2 days leave.
(f) 2 Way /Factorial ANOVA
Because there are 2 levels of Factor A (normal and deprived) and 3 levels of Factor B (0,2,4 cups of coffee) and we are finding a relationship.
(g) Independent sample t test
(i) there are 2 different populations, who are not related to each other
(ii) t test because the population sizes are < 30
(h) Single sample z test.
Only 1 group of people are there and they are being tested against the population average, with a known standard deviation.