INF 397C Fall, 2003 Days 13

Designing Research Finishing the exercise from last week . . . Question 3 (groups 5 and 6): Does design of a Web site affect visitors’ perception of the validity of the content?

Homework http://www.environ.wa.gov.au/community/results.asp http://www.yogapoint.com/info/research.htm http://www.main.nc.us/bcsc/Chess_Research_Study_I.htm http://www.econ.upenn.edu/Centers/CARESS/CARESSpdf/00-15.pdf http://www.georgetowncollege.edu/departments/education/portfolios/Franzen/research_paper.htm http://www.pbats.com/articles/foot/Foot_Postures.pdf http://www.cwrl.utexas.edu/currents/spring02/wakefield.html http://www.neuhaus.org/paper3.htm

Other “T scores” http://www.rehabinfo.net/rrtc/publications/research_summaries/neropsychfunction/default.asp “T scores” – “Standardized T scores for the WAIS-R, WISC-R, and HRB” http://collection.nlc-bnc.ca/100/201/300/cdn_medical_association/cim/vol-21/issue-2/0094.htm Yet new “T scores.”

Other measures Confidence intervals http://www.hortnet.co.nz/publications/nzpps/proceedings/95/95_294.htm Anova http://www.ischool.utexas.edu/~adillon/publications/empirical.html

Chi-squared Hinton, p. 239

Chi square test Let’s work an example. Just know that you use the chi square test when you have FREQUENCY data.

Correlation With correlation, we return to DESCRIPTIVE statistics. (This is counterintuitive. To me.) We are describing the strength and direction of the relationship between two variables. And how much one variable predicts the other.

Hinton, p. 259

Break

Let’s conduct some REAL research Research question: Is one of the two formulae for Standard Deviation easier to use? H0: µ1 = µ2 HA: µ1 ≠ µ2 Population: iSchool Research Methods students Sampling technique: Random IV: Particular formula (derivative formula vs. computational formula DVs: 1) Speed, 2) Accuracy

Procedure Who has a calculator? Number off, 1 to N. We’re gonna choose 10. Randomly assign 5 to Group 1 and 5 to Group 2. http://www.random.org/sform.html Group 1 – Derivative formula first. Group 2 – Computational formula first. For each testee (wink), need one volunteer to time the completion. Test participant: Please start when I say “Go,” and tell your observer when you think you have the answer.

Test Materials -- A Please calculate the Standard Deviation of the following distribution of scores: 5 6 2 8 3 9 4

Test Materials -- B Please calculate the Standard Deviation of the following distribution of scores: 4 6 8 7 3 9 2

Scoring Observers, please collect the following data: Test Participant number Materials A problem: Correct (y/n) Materials A problem: Time to complete? Materials B problem: Correct (y/n) Materials B problem: Time to complete? Difference score: Time for derivative (whether it was A or B) minus time for computational formula (whether it was A or B) Let’s build a spreadsheet with the data.

End game Calculate t. Reject or not? Conclusion. Now, REALLY, to whom do these results generalize?

Let’s talk about the final Course evaluation Q-and-A session, Monday, December 1, 5:30 – 6:30, place TBD. Sample problems handed out next week.