1.) How does a within-subjects experiment differ from a between-subjects experiment? The main difference is that in a within-subjects experiment each subject receive the treatment conditions instead of with a between-subjects experiment where different groups of people receive the treatments.
2.) Discuss three advantages and three disadvantages of using a within-subjects design. The three advantages are that you use less subjects, it’s clearer to see the effects of the IV, it controls for extraneous variables. The disadvantages is that it may take more time, experiments become tedious for the subjects, confounds can still occur.
3.) Outline a within-subjects experiment to test this hypothesis: Children who are given weapon-like toys become more aggressive. Have each child play with a normal toy and then rate their aggressiveness then have each child play with a weapon toy and measure their aggressiveness on the same scale.
4.) Mary is very excited about the within-subjects approach. “Now I’ll never need to run large numbers of subjects again,” she says. What has she forgotten? You’ll still have to run a large number and combination of treatments.
5.) For each of the follow dependent measures, evaluate the pros and cons of using a within-subjects approach:
a.) The taste of new toothpaste. Will use less subjects but the taste of the different toothpastes will produce a carryover effect.
b.) The cavity-preventing properties of a new toothpaste. Controls for extraneous variables like brushing habits but it would be easier to just have a control group.
c.) The readability of a new typeface. Will control for extraneous variables but will have to recaliber typeface after every use.
d.) The impact of good and bad news. Use less subjects but there could be potential order effects.
6.) What requirements must be met to make the within-subjects approach feasible? Need to control for order, fatigue, and practice effects. Must also be measuring something where subjects can be tested twice.
7.) You want to test whether different scramble patterns lead to different solution rates. You want to use the same words in all conditions so that the type of word will not be confounding variables.
a.) If you use a within-subjects design for this experiment, will you have to worry about order effects? Why or why not? Yes, because subjects will get practice solving scramble word problems.
b.)Review the four counterbalancing techniques for handling order effects. Which would help you most in this experiment? Why? I’d use partial counterbalancing because you will have six or more scrambles for each word.
c.) What are carryover effects? Would they be a problem in this experiment? How would you handle them? A carryover effect is when a treatment effects the one that follows it. This would happen in this experiment because if scrambles are similar than people will be able to solve the scramble that comes after the similar one faster. You would have to use Latin square counterbalancing.
8.) Studied the effect of audience laughter on the perceived funniness of recorded jokes. Subjects received four possible manipulations and judge how funny they believed the audience found the comedy routine. Subjects rated the comedy routine as less funny when they were asked this question first than when it came fourth.
a.) Is this experiment confounded by order? Yes because order influences the effect of the IV on the DV.
b.) Explain why or why not? Subjects can’t tell if an audience’s laughter should be considered extra funny or not because they have nothing to compare it to when it’s presented first.
9.) Commercial showed people tasting and choosing between two colas. One was labeled R and the other was Q. Majority of people said they liked R better than Q. Would you accept that ad’s claim that cola R taste better than Q? Why or why not? How might you change the procedures to get more acceptable data? I wouldn’t accept the claim because the taste of one cola may affect the taste of the one that comes after it. I would change the procedures to reverse counterbalancing.
10.) Explain why (when it is used as a factor in a design) order is always a within-subjects factor. Because order can affect your data when you are dealing with the same person. If you have two groups order isn’t a factor because each group is getting a single treatment.
11.) Based on figure 11-6 what strategy would you recommend for handling progressive error in this experiment and why? I would use a randomized partial counterbalancing and make sure there was more time between when they tasted each candy.