The primary reason for using a repeated measures design rather than an independent groups design is that it is unethical to have a control group where juvenile human subjects participate and some do not receive appropriate treatment. In the case where one treatment is reasonably judged to be effective, that treatment cannot be withheld merely to create a control group.
Another reason is that repeated measures design is more useful because it controls for variability among test subjects, and it can directly relate specific outcomes to specific individuals, i.e. the design allows the author to directly observe improvements, or lack thereof, in each individual independently. Another reason was that the small control group in the study was simply not enough for quantitative analysis, which would not have been very useful anyway, so the qualitative analysis was done for each individual case and the conclusions were drawn from similarities and differences.
Many researchers agree that in psychology, qualitative analysis often provides more insight into the subject of study than mere quantitative analysis. After all, it is really quite impossible to create a true control group of human subjects, since no two individuals are alike. (Unesco 548) More useful inferences can be gleaned from anecdotal evidence than from mere numbers if the evidence is truly thoroughly described and carefully examined.
The effectiveness of the treatment was assessed in a six-month follow-up and compared to results from the pre-treatment assessment. The results indicated that the treatment was definitely effective in reducing aggressive verbal behavior, and in increasing the use of verbal reasoning and communication, but the treatment didn’t significantly affect the expression of physical aggression. This might indicate that this kind of treatment can affect thought or cognitive processes, though emotion may remain high even so, a factor that impacts physical aggression. In addition, a teacher assessment indicated that the subjects’ behavior was significantly improved following treatment. Therefore the treatment did appear to be, at least partially, effective in improving overall behavior.
Noting that the treatment seemed to reduce verbal aggression, though not necessarily physical aggression, points to the difference in these modes of expression of emotion. Often physical aggression is brought on by emotional factors. In fact, sports coaches use 6this to whip up the team before the competition.
From Wikipedia: “Internal validity is a form of experimental validity. An experiment is said to possess internal validity if it properly demonstrates a causal relation between two variables. An experiment can demonstrate a causal relation by satisfying three criteria:
- the “cause” precedes the “effect” in time (temporal precedence),
- the “cause” and the “effect” are related (covariation), and
- there are no plausible alternative explanations for the observed covariation (or nonspuriousness). “ (Shaddish, Cook and Campbell 2002)
Some of the threats to internal validity in this study:
It is possible that students would have naturally shown reduced levels of aggressive behavior and overall improvements without the treatment of this study over time. The study assumes overall dysfunction in the students would have remained the same over time, but this is neither certain nor even probable. Even the author mentioned this factor and noted that few children suffer very long-term effects from divorce, and some who do so are affected despite treatment.
“Despite such resilience, many children of divorce experience intense, short-term effects that negatively impact upon their school performance; and a few children carry long-lasting effects into their own adulthood that seriously impair their ability to develop and maintain a long-lasting relationship.” (Wallerstein, 1988; Wallerstein & Corbin, 1999 in Dykeman 2002)
Children are often resilient and they recover better than adults in situations such as these. For example, many students might ‘act out’ immediately following their parents divorcing, but even without treatment would naturally ‘settle down’ over time. (this violates criteria 3) Because there is no control and due to the individuality of each situation and subject, there is no way of knowing what might have resulted if nothing was done. When dealing with human subjects, it must be noted that all evidence is, necessarily, anecdotal and analysis must be carefully reasoned.
Participant expectations could also have influenced the outcome- i.e. since the troubled students were in treatment, everyone involved, including the students, probably had some expectations of improvement. This alone can produce an effect. In other words, the effect seen could have been equivalent to a placebo effect. (this violates criteria 3) Professor Ernzt reminds us that the term ‘placebo response’ includes and thereby confuses and obscures a host of ill-understood effects; effects that influence all forms of therapy to a greater or lesser degree. Improving certain aspects of the therapeutic encounter may enhance the outcome of any treatment, so Professor Ernzt encourages practitioners to learn about these elements and to use them well in everyday practice.
He reminds us that the phrase so often used about complementary medicines, that they are ‘just placebo’, typifies the denial and misunderstanding of the many non-specific effects that contribute to self-healing responses and therapeutic outcomes. “Since 1955, when H. K. Beecher published his classic ‘The powerful placebo’, it has been generally accepted that 35% of patients with any of a wide variety of disorders can be treated with placebos alone.” (Kienle, and Kiene 31)
However, in another sense, it is possible that since many psychologists expect that about 30% of problems may be successfully treated with the placebo, the internal validity of this study is only affected to the extent that the treatment, itself, is deemed responsible for positive or negative changes. Perhaps using a bell curve with the top and bottom 10%-15% discounted would eliminate some of this uncertainty if that is deemed necessary.
There was no significant reduction in physical aggression. This would be considered by some to be the ultimate test of the success of treatment, and it was missing entirely. It can be argued that a causal effect has not been properly established. (this violates criteria 2) However, if we consider that verbal aggression and physical aggression may have different causes and effects, we may need to take another closer look at the complicated psychological theory being applied here.
In studies of children and violence on television, many researchers have found that the effects of violence on physical aggression are short-term. It is thought that verbal aggression and physical aggression do not have the same causal factors, nor the same resolution. “Lynch and Cicchetti’s (1998) longitudinal study of 322 ethnically diverse children from disadvantaged backgrounds found that the relationship between ETV and violent behavior became nonsignificant after controlling for prior behavioral functioning (i.e., violent behavior).” (Spano, Rivera, and Bolland) Such behavior stems from different cognitive factors.
One interesting quote from the article pointed out the different factors that the researchers were aiming at in the treatment: “The treatment of the cognitive, affective and behavioral consequences of divorce at the community agency can extend across the continuum of primary, secondary and tertiary intervention. (James & Gilliland, 2001).” (Dykeman 2002) In other words, these researchers were trying for a “shotgun effect”, by aiming therapy at three different areas of responses: cognitive, affective, and behavioral.
IN looking at behavioral psychology, we must note that Skinner (1932) correctly predicted that behavior has a strong effect upon the psychological feelings of the subject. That is, smiling can make one feel happier, and crying may do the reverse. Therefore, the “so-called placebo effect” in psychology may be a valid mechanism for behavioral change. This study showed that for certain types of aggression and poor behavior, the intervention had certain positive effects, That is the nature of psychological research: cause and effect may be deduced and studies, but the exact mechanisms at work in each individual may not be the most obvious.
References
Shadish, W., Cook, T., and Campbell, D. (2002). Experimental and Quasi-Experimental Designs for Generilized Causal Inference Boston:Houghton Mifflin “On the rate of formation of a conditioned reflex”. Journal of General Psychology, 1932, 7, 274 -86.
Kienle, Gunver S., and Helmut Kiene. “3 A Critical Reanalysis of the Concept, Magnitude and Existence of Placebo Effects.” Understanding the Placebo Effect in Complementary Medicine: Theory, Practice, and Research. Ed. David Peters. Edinburgh: Churchill Livingstone, 2001. 31-50. Questia. Web.
Unesco. Contemporary Political Science: A Survey of Methods, Research, and Teaching. Paris: Unesco Publication, 1950. Questia. Web.
Spano, Richard, Craig Rivera, and John Bolland. “The Impact of Timing of Exposure to Violence on Violent Behavior in a High Poverty Sample of Inner City African American Youth.” Journal of Youth and Adolescence 35.5 (2006): 681+. Questia. Web.