The aim of this paper is to give a review of the article “Training Forgetting Negative Material in Depression” written by Joormann et al. (2009) by providing an analysis of the effectiveness of the article’s style in conveying the research’s importance and contribution to the study of depression as well as the appropriateness of the experimenters’ approach to conducting their study. Briefly, Joormann et al.’s experiment investigated whether using substitution as a specific cognitive strategy can aid depressed participants to forget negative target words. Using the think/no-think method, they found that by substituting unwanted target words with another, clinically depressed participants are able to forget a significant number of the words they learned in an earlier phase. These results supported the idea it is possible to train depressed individuals to counteract their rumination tendencies.
Upon reading the title of the article, readers are able to grasp the general idea of the experiment. One can deduce that the experimenters are intentionally trying to make depressed individuals forget negative materials by training them. Although the title makes sense, it is nevertheless, slightly misleading. The word “material” is a very broad term and can refer to a number of meanings – physical matters from which an object can be made, things needed in an activity, facts and information. In the experiment, however, only words were used. The word “material” in the title only refers to words and not any other sort of medium that can be considered a material. While trying to generalize their findings beyond the scope of the experiment, the writers of the article can be said to have overgeneralized from “words” to “materials”. This title may mislead the reader to expect more in the experimental design than what was actually presented. A more suitable title may be something like “Cognitive Strategies Aid Forgetting of Unwanted Words in Depression.”
Both Abstract and Introduction give sufficient information to understand what is going on with the present study, with goals clearly stated. The writing style is professional and scientific. The subsequent Method section is less forthright and requires more explanations. Because the procedure was rather complicated and required advance material preparation, the explanations could be better clarified if diagrams were available to show the different experimental conditions and the manner of which the word pairs were created. Also, some paragraphs may be read more smoothly had there been tables showing the many statistics instead of having them disrupt the sentence flow, such as the paragraphs explaining the final sample and design, which were engrossed with statistical data.
In Results, there were several instances where the choice in words was rather unsuitable. Stressing the idea that the number of cue presentations influenced the number of words participants recalled, Joormann et al. wrote “…conditions depended on…,” “targets depended significantly on…,” and “…control group depended upon…” (italics added, p. 38). There was a strong correlation between conditions and the amount of cues participants were presented, but this certainly does not imply that one variable controlled the other in such a way to create dependency. Less causal words should be used. Additionally, the two graphs presented in this section could have been better presented. The error bars in both Figures 1 and 2 are poorly indicated; one can hardly differentiate error bars of one condition from those of another without scrutinizing the graphs to figure it out. Three small separate graphs juxtaposed could have made the reading easier.
Due to the many conditions in this experiment, the actual number of participants per condition is actually not that many. A total of 90 participants carried through from the beginning to the end of the experiment, but there were only 15 per condition. (There were 45 participants each in the control and MDD group, with three conditions per group.) Calculating statistics with only 15 sets per condition allows room for swaying the results, driving it to a significant level. More participants should be recruited to resolve this problem.
In the Methods section, Joormann et al. wrote that participants only needed to recall 50% of the target words accurately to go on to the TNT phase. This is a surprisingly low passing criterion. Because passing scores were not disclosed, it is hard to judge whether the majority of the participants correctly recalled significantly more than 50% of the targets or barely passed. This confusion may be solved if the data was made clear. Suppose this low criterion is due to time constraints, so participants were not expected to remember the majority of the word pairs. Understandably, the pass criterion is set low. However, this certainly impacts whether the findings of this experiment are generalizable beyond the laboratory. Assuming words are representing the thoughts that constantly brood in the minds of depressed people, these words would not be so easily forgotten due to constant rumination. A low criterion means that participants may pass even if they barely remember half of the targets and that variation in the final recall test would vary greatly. Moreover, this ties in to the baseline condition; many word pairs were not retained in the learning phase, which resulted in low recall during the recall phase. This affects the baseline condition, although the actual baseline statistics were not made explicit.
The Discussion section provided a good wrap-up of the article. It reiterated what the study’s purpose was and how they approached investigating it. Yet, the writers concentrated a great deal on connecting their present study to past studies. Further, different background information, studies that explored cognitive biases in attention and stress reactivity, were considerably elaborated. The background help readers acquire a better grasp of this line of research, but it is, nevertheless, not necessary to provide so much details on what this study is not analyzing. The discussion should have emphasized explaining why both positive and negative substitutes helped depressed participants forget and why it is best to use the TNT method rather than other distraction methods to aid forgetting. Also, one major limitation the researchers mentioned was the unclear underlying mechanism behind the results they acquired. A simple recognition test of all the target words should have been administered. This would show whether participants in both conditions completely forgot or did not encode the word pairs or whether it was just an inability to retrieve information because of competing words. If it was the latter, seeing the target words again may trigger word retrieval and the results of this recognition test would show that targets have not forgotten the materials they learned.
Joormann et al.’s article is logically organized but lacks clarity in some of the more important sections. The graphs could have been better presented and charts of statistics would have helped the article flow better. The writers emphasize the incompleteness of their research and digress on providing more background information rather than discussing the importance of their findings. Overall, readers get a good idea of what is being done in the study but at the same time, may be lead to confusion due to the complications of the procedures.
Joormann, J., Hertel, P., LeMoult, J., Gotlib, I. (2009). Training Forgetting of Negative Material in Depression. Journal of Abnormal Psychology, 118 (1), 34-43.