Posted: March 11th, 2023

Discussion Three

Discussion Three

This is one of the most important parts of this course – I want you to work with your group to design a follow-up study using the topic of the Self-Reference Effect, or SRE. Your instructor will present the ideas to the whole class, and you will vote on which one you would like to do for your final class project. As you work on this discussion, try to think about what other variables might influence SRE. Let me give you a few examples, though try to get creative with your group. Look at prior research and see if you can use some of their work as a follow-up to your study. Have fun with it!

Okay, consider some ideas I was thinking about for a follow-up study (Personally, I would recommend the third or fourth idea, as there is a lot of research you could draw on these new independent variables, which would make Paper III much easier to write). Keep in mind that we will drop one condition from our original Study One independent variable. I suggest keeping the Self-Rating and Friend-Rating conditions and dropping the President-Rating condition. I have no solid reasoning behind this other than the Friend-Rating and President-Rating conditions probably will not differ much in study one, so including both is not necessary. I think keeping the Friend-Rating condition will be easier to use with our new study design for some of the ideas below, but we could retain the President-Rating condition if you have a good reason for doing so!

First, if we keep the Self-Rating (SR) and Friend-Rating (FR) conditions, we could alter the valence of our descriptive terms. That is, for some participants, “Social Media Post #1” only includes negative words (e.g. annoying, irresponsible, etc.) while others only see positive words (e.g. likeable, friendly, etc.). Would participants recall more words in the positive word condition than in the negative word condition? (Note that we would have FOUR conditions in this design: SR and positive words, SR and negative words, FR and positive words, and FR and negative words). If you think this is a good idea, look at some research on memory for valence or emotionality!

Second, we could alter whether we instruct participants to think about whether a word DOES describe versus DOES NOT describe either themselves or their friend. Again, we have four conditions here (SR and does describe, SR and does not describe, FR and does describe, and FR and does not describe). I suspect it may not matter. Whether you decide if something does or does not describe yourself, you are still thinking about how the word applies to you. But it might be interesting to test out this possibility.

A third idea might involve asking participants to think deeply about themselves before seeing “Social Media Post #1”. For some, we ask them to think about their friends and what makes them unique among their friends (unique condition). For others, we ask them to think about what similarities they have with their friends (similarity condition). They then do the study similar to study one, with some in the Self-Rating condition and others in the Friend-Rating condition. How might thinking unique versus similarity thoughts about themselves impact participant recall of the social media post words later in Part A (Recall)? It is possible that the “thinking” exercise might get rid of the self-reference effect for everyone, even for those in the FR condition. But again, I might be wrong!

A fourth idea, and one I think is interesting, is adding in a “primacy effect” idea by having the “Social Media Post #1” words start positively and end negatively for some participants (positive condition) but start negatively and end positively for other participants (negative condition). Will recall differ if the valence of the words start positively versus negatively? It might be interesting to explore this possibility. If the “primacy effect” works (and idea that items at the start of a list are better remembered than items at the end of the list), the SR participants might recall more positive words in the positive condition and more negative words in the negative condition with no differences for FR participants.

Finally, you can look at some participant characteristics if you like. You can measure the participants’ familiarity with similar “personality quizzes” on social media. Those who like to do these tests online might respond differently than those who ignore such online personality quizzes. Or we could also look at participant gender. Are men or women more likely to engage in self-referencing? Just remember that if you choose a participant characteristic as your second independent variable, you cannot draw cause-effect conclusions (since you cannot randomly assign someone to be familiar versus unfamiliar with online personality tests, and you cannot assign someone to be a male or a female). That is why I prefer that you manipulate some feature of the study so you can randomly assign people to one of the four different conditions.

As you can see, there are tons of ways to extend your original study. Work with your classmates to identify the one you like best, and then let your instructor know. Your instructor will take the consensus vote for the whole class and let you know what the study will be for study two. In this discussion, I want EACH of you to do three things.

First, tell me which study you want to do (that is, which second independent variable you find most interesting). You can use one of the ideas above or come up with a unique one on your own. I prefer your own ideas, but if you really like one I mentioned you can use it. You still need to give me two more things, though …

Second, give me a reference in APA format for one peer reviewed research article that has something to do with this second variable. This article 
does not have to involve Self-Referencing at all, but it must have something to do with your second independent variable.

Third, give me a hypothesis for what you expect to occur if your new independent variable is chosen for the class project. What do you predict just for that independent variable? What do you predict in terms of the interaction of that new independent variable with your original Self-Reference Effect (Self-Rating vs Friend-Rating) independent variable?