You may want to discuss or assign the following article:
Hemenover, S. H. (2003). The good, the bad, and the healthy: Impacts of emotional
disclosure of trauma on resilient self-concept and psychological distress.
Personality and Social Psychology Bulletin, 29, 1236-1244.
The author uses a mixed design to study an interesting topic (the benefits of writing about traumatic experiences). In addition, the article is easy for your students to obtain (students who buy the book can get it by using the Infotrac® subscription that comes with Research Design Explained), and the article is only moderately difficult for students to read (to make it less difficult, give students Table 1).
The author explores whether writing about a traumatic event—an activity that has beneficial effects on physical health— can have beneficial effects on psychological health. He provides several reasons to expect that writing about traumatic experiences may help repair —and possibly enhance—one’s self-concept.
On three separate days, participants came to the lab and were asked to write for about 20 min on a neutral topic (tomorrow’s plans) or about a traumatic life event. For each writing sample, a computer program counted how many positive emotion words, how many negative emotion words, how many insight words “(e.g., think, know)” and causation words “(e.g., because, effect)” participants wrote.
Before doing their writing assignments, all participants were administered a pretest consisting of a multidimensional measure of resilient self-concept (Ryff’s [1989]Scales of Psychological Well-Being) as well as a multidimensional measure of psychological distress (the Revised Symptom Checklist [Derogatis, 1977]). Three months after doing their writing assignments, participants were administered a posttest consisting of those same two measures.
A manipulation check showed that participants in the trauma-disclosure condition “used more negative emotion words and fewer positive emotion words” than controls. In addition, trauma participants viewed “their narratives as more upsetting…than did control participants.” Using “a series of 2 (condition: trauma vs. control) X 2 (time: pretest vs posttest) repeated measures MANOVAs,” the author found that the trauma participants changed more from pretest to posttest than the controls did. The author argues that “among trauma participants, the greater the increase in their use of insight words over the three writing sessions, the higher their autonomy and the lower their interpersonal sensitivity at posttest.” However, analyses of the writing samples did not lead to any clear evidence about what mediated the effect. The discussion focuses on the value of writing about trauma, the possibility that even the small effects observed in this study may be very important and long lasting, the need to find mediators of the effect, and the need to study this phenomena using a noncollege student sample.
Helping Students Understand the Article |
|
|
Section |
Tips, Comments, and Problem Areas |
|
Abstract |
“trauma participants” refers to participants who wrote about a traumatic event. “resilience”—as you will see later in the article— is indicated by feeling competent, feeling in control, wanting to continue to improve, caring about others, seeing one’s life as having meaningful life goals, and accepting oneself for who one is. |
|
Introduction |
“salutary”: healthy, beneficial “content analysis”: for more about, see pages 152 and 196 of your text. “coherent”: consistent, organized “disparate”: different, inconsistent “self-regulation”: able to control one’s own thoughts, feelings, and actions “self-empathy”: self-understanding “efficacy”: being effective “obsessive ruminative”: lengthy and repetitive “bereavement”: grieving “parameters”: aspects of “moderate”: alter (for more see, pages 51 and 52 of your text) |
|
Method |
“SD” is an abbreviation for standard deviation (a measure of how spread out the scores are) “somatization” the degree to which emotional problems are expressed in the form of physical symptoms. “all as >.80” indicates that each subscales’ items were internally consistent (see pp. 101-104 in the text). “… manipulation checks …” If control participants had used fewer positive emotion words and more negative emotion words than the participants who were supposed to be writing about traumatic experiences, the researchers would have suspected that participants were not taking the task of writing about traumatic experiences seriously. |
|
Results: Manipulation check |
“inflation of Type 1 errors”: having a high risk of stating that there is a relationship when there is not, especially compared to the risk you claim to be taking (see pages 174-175 and page 212 in your text). “MANOVA”: multivariate analysis of variance (see page 212), when you have several (multiple) scores for each participant, doing one MANOVA is an alternative to doing multiple ANOVAs. “univariate tests”: tests that look at scores on one (“uni”) dependent variable (“variate”) at a time—in this case, an ANOVA. “h2 ”: a measure of effect size The last paragraph of this section provides a nonstatistical summary of the results. |
|
Results: Effects of writing condition on resilient self-perception and psychological distress |
“guard against Type 1 error”: The risk of making a Type 1 error (declaring that at least part of the difference between conditions was due to the treatment when it was really all due to chance) when using an ANOVA to test the effect of the treatment on a single measure would usually be about 5%. However, the researcher had multiple measures and thus would have to do many ANOVAs. The chances of one of those ANOVAs being significant by chance alone would be much higher than 5% (to see why doing multiple tests inflates the Type 1 error rate, see page 308). “Condition X Time interactions” indicate that participants who write about the trauma change more from pretest to posttest than participants in the control condition. To get a better idea of the differences in how the two groups change from pretest to posttest, go to Table 2 and look at the column labeled “D (Posttest-Pretest). Asterisks in that column indicate a significant difference between pretest and posttest. As you can see, there is only one asterisk in that column that relates to the control condition—for autonomy. Control participants scored significantly lower on autonomy (feelings of being independent and in control of one’s life) on the posttest than on the pretest. In other words, they get worse on this dimension. On the other dimensions, control participants did not change significantly from pretest to posttest. As you can see, most differences are near 0 (e.g., .06, -.03, etc.). Participants in the disclosure condition, on the other hand, changed on most measures. |
|
Results: Language analyses |
“insight words”: words like “think” “causation words”: words like “because” “insight3 – insight1” subtracting number of “insight words” used in first writing session from the number used in the third session. “three-step hierarchical multiple regression models.”: Conceptually, their procedure is very similar to doing an ANOVA (specifically, a 2 [time: pretest vs. posttest] X 2 ([condition: disclosure vs. control] ANOVA. The first part of this section is summarized by Table 3 (and by the last two sentences of the second paragraph). “dummy coding”: arbitrarily assigning a number to the different values of a variable (e.g., 0 = male, 1= female) so that the regression equation (which requires numbers) can use that variable. Note, that the number is arbitrary: We could have used 0=female, 1= male or 1=male, 2=female, or … |
|
Discussion |
“salubrious”: good “interpersonal sensitivity”: the author means feeling inferior to others (less interpersonal sensitivity is good; more interpersonal sensitivity is bad). “extant”: existing “self-efficacy”: feeling capable of doing a certain task “hardiness”: ability to recover from trauma “reciprocals”: opposites in the sense that the more you have of one quality the less you will have of the other quality. “impinges”: hurts, limits “ruminate”: think about something for a long time |
Back to Featured Articles Main Menu
Back to Having Students Report on a Within-Subjects Experiment