History as the Mediator of Meta-Analysis

In regression class I finally understood the difference between mediation & moderation.  A significant break through in the way I thought of variables influencing one another.  The most interesting thought was the relationship between time & meta-analyses. 

During a meta analysis we collect as many studies as possible which experimented on or evaluated  a particular variable.  We then compile these studies, do some rather interesting statistics and postulate a conclusion about the influence of this  particular variable on a dependant variable.  An example would be taking all the studies on Self-Esteem and exploring its influence on the personality trait extroversion.  However, in my work with the psycho-market model, slopes & relationships gain & lose significance over time.  A relationship may be significant one year and non-significant the next.  Time has mediated (or something else) the significance of this relationship.  In running experiments we try to control for "the something else," but we often never control for time.  I propose that the fluctuation of effect sizes & significant/non-significant relationships are mediated by not only standard experimental issues, but time.  I would go as far as saying that time would account for the majority of variance above & beyond standard experimental issues related to the fluctuation of meta analysis data.

Several time & other related errors do occur that need to be addressed.  When conducting a psychological experiment we often forget it was done at a certain period in time, with a particular zeitgeist.  Therefore when conducting meta-analyses we may be combining significant & non-significant time sensitive data which may eliminate any true relationship or produce a type 1 error.  In addition, every experimenter knows of the bias journals have towards significant results & meta-analyses typically only use these results, collecting very few datas from non-significant sources.  Lastly, it's erroneous to think that the knowledge an individual holds about Self-Esteem or Organizational Justice was the same when the study was first conducted to when the meta-analysis was calculated.  Some erudite professors may rebuttal with "This is why we operationalize constructs."  My response to that would be look at the literature on survey responses & actual occurrences of these events. I think one would be hard pressed to find someone who hadn't incorporated their own version of what the variable means.

In short, I would attest that the fluctuation across experiments, evaluating the same variable, is most attributable to time above & beyond standard experimental errors.  The question becomes how can we control for time?  I have outlined some suggestions.

1) Use a linear variable to depict the change of time.

2) Create a zeitgeist construct to be included in the analysis

3) Understanding this may be one variable that is uncontrollable & accepting it's placement within the experimental meta-analytic frame work.  Therefore, not accepting meta-analyses as the end all tell all, as they often are.

Equity Traders & Attribution Theory

Attribution theory, according to Wikipedia (which offers a very accurate definition) defines it as "The theory that is concerned with the ways in which people explain (or attribute) the behavior of others or themselves (self-attribution) with something else." Attribution theory was originally pioneered by Fritz Heider in his conjecture that humans operate in amateur scientist ways; mainly that we gather information and create internal/external causal inferences.  

Research in this area began to unravel the complex mechanisms in which we humans make these inferences.  Most important to us as investors are the mechanics that operate when we are 'depressed' or 'happy'.  Individuals use three grand schemeas to categorize attribunal inferences 1) Internal/External, 2)Stability, 3) Control.  Specifically for depressed  individuals, they attribute failures to: Internal, Stable, and Uncontrollable properties.  An example would be, "I can never pass this test, because I'm dumb and will always be dumb."  Very extreme but makes the point.  In the instance of a success: External, unstable,  and uncontrollable.  An example would be, "I got this A because the teacher likes me and I was lucky."  On the Flip side individuals who are happy respond in a polar way.  For failure attributions are: External, unstable, and uncontrollable.  For success attributions are: Internal, Stable, and controllable.  

It's a very big difference.  The most significant difference to note is that while depressive individuals blame themselves for failure, 'happy' individuals do not.  In the instance of success depressed individuals attribute it to something other than the self while 'happy' individuals take the credit.  On one hand it's very self defeating and on the other very self protective.

How does this relate to traders & How can it help us?  Very simply actually.  Using the Psycho Market Model we need to look for the ways traders attribute their failures & successes during a span of trading time (days, weeks, years, etc).  Are traders looking to blame the market, their system, themselves, or something else?  If we can get at the core of traders attribution styles we can use the plethora of information from entity & learning theory to 'reteach' traders how to make accurate attributions regarding their trading activity.  Accurate attributions will allow traders (if they so choose) to introspect and become meta analytic machines which can recognise their own weakness because they know where the fault lies.  Turning those weaknesses into large profits.

A very interesting variable that does not exist in life are stop loses.  How does this variable interact with attributions of success or failure in trading equities?  Research needs to be conducted in this area to enlighten market participants of their own possible short comings.  

In the coming months I will be developing a traders version of the Attribution Styles Questionnaire originally developed by: Peterson, C., Semmel, A., von Baeyer, C., Abramson, Metalsky G.I., & Seligman, M.E.P. in 1982.

Credit where credit is due:

Dan Ariely  had suggested that if I were to pursue a model that synthesized economics & psychology that I may begin by looking at how traders may vary from 'normal' people.  This is one instance, in which they may.  

References:

Hewitt, A., Foxcroft, D., & MacDonald, J. (2004, November). Multitrait-multimethod confirmatory factor analysis of the Attributional Style Questionnaire. Personality and Individual Differences, 37(7), 1483-1491.

Peterson, C., Semmel, A., von Baeyer, C., Abramson, L.T., Metalsky, G.I., and Seligman, M.E.P. (1982). The Attributional Style Questionnaire. Cognitive Therapy and Research, 6, 287-300.

Attribution theory. (2009, March 18). In Wikipedia, The Free Encyclopedia. Retrieved 22:54, March 18, 2009, from http://en.wikipedia.org/w/index.php?title=Attribution_theory&oldid=278139181

My notes from Dr. Johnson's social psychology class at Hofstra University

Growth is Truely Exponential!

After being rejected by seven I/O PhD programs & two quantitative psychology programs, needless to say, my spirits were low. I desperately tried to make a case for synthesizing psychology & economics as a course of study in these programs. I tried to explain the relevance and the potential impact it may have upon the field of psychology. My words feel upon deaf ears. A retired professor I spoke with, who was the creator of market distribution channels as organized systems, said it best "If you want to make it in academics you have to plow what has already been plowed, just plow it deeper and more scholarly." I've never heard it put more candidly than that and realized that this would not be an easy journey. I began emailing professors again asking for advice, how they started their careers etc. Economists told me it was to psychological and psychologists told me it was to much economics. I spoke with names such as Dan Ariely, Bruno Frey, Alois Stutzer, and Fiona Lee to name a few. The search ended with little to show.

At last though I see some light! The old clinical gears in my head swirl with anticipation! Proliferation is in sight! I recently contacted Philip Pearlman who kindly spoke with me about his ideas regarding the market & psychology. I found someone who I agree with and I think will agree with my ideas as well. I may finally be able to bring the psycho-market model into fruition. In addition, my first Psycho-Market Model Oil Report, taking into account Psychological & Economic factors is coming to completion as well. Also, a PhD student Sayeed & I are going to begin a line of research regarding President Obama's Speeches. Additionally, my consulting work with a Rhode Island Bed & Breakfast, The Spring Seasons Inn has reached its first stage of completion. The analysis of past performance is complete and we will begin the development of customer surveys shortly. Lastly, I'll be graduating with a Masters Degree in Industrial Organizational Psychology from Hosftra University in May.

After much frustration & failure it's nice to stand up and let the scabs heal for a bit. Though I'm not afraid to get bruised again down the journey I am about to travel through. Growth for me has happen so quickly in such a short period of time that it seems unbelievable. Looking back at how miserable I felt at times it all seems worth it and is well noted to keep this in mind when experiencing even more difficult hardships.