Hello. I have to admit that I haven't directly encountered that categorization of aggression, however there are some beginning steps that you need to take. First, have you done a very thorough search for any existing measures? It seems fairly common today that people modify existing questionnaires, so even if you found something similar it might help you. Also I've been very lucky when I've been looking when I use dissertations. Given I lack information about your chosen methodology, I am going to stick to general points.

If you are going to use self or other report then try to find a valid and reliable questionnaire to use. Behavioral observation is also extraordinarily useful, and is more flexible when you design the observation versus creating a measure from scratch and then having to go through the process of validation. Or, conducting a fairly rigorous study with a good sample and estimated power THEN rely on self or other reports. Having worked in a federally funded clinic for years that used self-report, other report, and behavioral report we noticed a few things. One, at the beginning it made sense to include all sources or as many as we could and to include measures that assessed (theoretically) cognitive processes of interest, but also to include behavioral observation. One of our students did his master's thesis on the correlations among self, other report and behavior observation and found that they were literally unrelated. Which wasn't a huge surprise as we had noticed certain patterns. However if you do use behavior observation, to my knowledge is simply take the existing definition or a reliable and valid, brief other report and take the items and convert them into behaviors that with some in-house testing you can find out fairly quickly whether these behaviors that you'll be training individuals to look for are identifiable, reliable, etc.. It's more the training of the typically students and the maintaining of a certain level of accuracy that is time intensive and absolutely critical.

Assuming you have your sample, IRB approval and you're basically ready to go then you probably have figured out that among your students, some are significantly better than others at behavior observation. As it turns out, we've known for a while terms like fear, aggression, acting out, are too vague and at this point its important that you spend enough space in the beginning of your article if you intend to publish defining this particular categorization and why its preferrable or not. This is necessary even if that's what you're doing this study to find out. Following closely behind would be a very explicit discussion of what behaviors, motivations, thoughts etc characterize the children in each category, again even if its theoretical at this point. You'll need to return to at least some of those issues in your conclusion.

Lastly and this is not widely agreed upon, I personally think that when you're doing what may appear to be clinical studies with children that you need to demonstrate knowledge about children of the age that you would be including in your study. For instance, in an article that was published years ago, the author was fairly firm that one particular measure used with children and adolescents, developed by our lab, and of course used in our studies, should not be used in any research until further clarification was obtained about its reliability and validity in terms of Piaget's work. luckily my developmental training had focused extensively on both Gibson and Piaget and I knew he had never said that children who were 12 and older could in fact answer, reason, about their own bodies and their feelings, but children who were age 11 and younger could not. Since our measure was being used with children ages 6 to 18 the contention was about half of the participants were unable to answer the questions accurately so they were simply telling us what we wanted to hear or giving random answers. Personally this article came out when I was just beginning to enjoy research so I very quickly gathered the information needed to show what Piaget actually ssaid about age and reasoning ability. My main grad student and life saver ran the stats, which indicated support for children roughly aged 8 and older having the capacity to reason in the manner that was required by the questionnaire. The editor wanted to be sure that we had done enough statistical work for a challenge such as this, so it ended up being reviewed by a clinician, a developmental psychologist, a statistician, and a methodologist. Again my grad student stepped up, learned some very new, at that time, methods to increase the likelihood that we had not made an error in our use of statistics. These new methods indicated our original statement held up and in combination with a very solid conceptual developmental framework were enough to convince the editor to publish. ~I'd like to say that anyone involved with even a slight overlap of clinical and developmental questions, using samples of individuals aged 18 and younger need to assess and include developmental ability as it relates to the study. I hope this wasn't too long and I hope it helped you.

Hi katy,

I suggest you the following paper:

Little T. D., Jones S. M., Henrich C. C., Hawley P. H. (2003). Disentangling the “whys” from the “whats” of aggressive behaviour. International Journal of Behavioral Development, 27(2), 122–133.