Learning Objectives
- Learn more about the statistical analysis tools inSPSS.
- Practice manual calculations and confirm with SPSSoutput.
- Walk through all steps of hypothesis testing for arelated-samples t-Test.
About Your Data
A statistics student was late on completing the final statisticsproject. While racking her brain for ideas, she watched her cateating supper and noticed how much quicker the cat ate the newbrand of cat food. Feeling desperate about the project, she decidedto figure out if cats really do prefer Brand A over Brand B. Shecatnapped six neighborhood cats (No judging!) and deprived them offood for 10 hours. Then she gave the cats Brand A, measured theamount of food eaten, deprived them of food for another 10 hoursand gave them Brand B. The amounts eaten in gramsare listed below:
Cat | Brand A | Brand B |
1 | 32 | 33 |
2 | 40 | 30 |
3 | 36 | 32 |
4 | 37 | 29 |
5 | 37 | 30 |
6 | 36 | 25 |
Instructions
The steps below are slightly different from the ones laid out byyour textbook author. I want you to follow my six steps below.Also, because it can be difficult and time consuming trying tofigure out how to insert symbols into a word processing document, Iwould like you to complete your manual work with pencil andpaper.
Six Steps of Hypo Testing and Manual Calculations
You will be conducting a related-samples t-Test (SPSS calls it aPaired-Samples test) to determine whether the cats have equalpreference for Brands A and B. Your alpha level will be .05.
Write out your six steps for this ethically dubious thoughhighly entertaining study. ;-)
1) State H0 using symbols AND words
2) State H1 using symbols AND words
3) State your alpha level
4) Articulate your rejection rule (\"Reject H0 if|obtained t| ≥ critical t, tcrit =_____\")
5) Manually calculate your related-samples t value (Showyour work. Carry your calculations out to two decimal places. Feelfree to calculate standard deviation on calculator.)
6) State your conclusions in four parts. a. Whatdid you do? b. What did you find? c. What does it mean? d. What canwe do with these findings? (REFER BACK TO ABOVE VIDEO forexplanation and examples of the different parts of Step 6. Followthe template provided in the video.)
Effect-Size Statistics
What are Cohen's d and the confidence interval ofthe difference? The confidence interval is listed in the SPSSoutput. You can do a quick internet search for \"Cohen's d\"calculator and find many websites that will allow you to quicklyperform that calculation. Once you have your effect sizes, explainwhat they mean.
Decision-Making Errors
Of course, there is always the possibility of Type I or Type IIerror. Which type of error is relevant in this case? What are somepossible consequences of this type of error?
