52 Weeks of Data Pattern Analysis: Week 4: What is Your Epistemology of Change?

What is Your Epistemology of Change?

Catching up?  The posts at http://www.datapatternanalysis.com/ have the previous week’s posts.  The activity section on my LinkedIn profile has them as well.  https://www.linkedin.com/in/thomas-arnold-phd/

This week, I will continue exploring the work I did with criminal offender risk scores.  Recall from Week 3, I had finished my Master’s thesis with a “paradigm conundrum.”  The Level of Service Inventory-Revised (LSI-R) was supposed to be able to measure changes in recidivism risk.  Based on the prevailing theory, my results opened up the possibility that the LSI-R could not measure changes in risk.

The average person might not have thought too much about this, but I had been thinking about this a lot.  See my post on the “slow change method.”

https://www.linkedin.com/feed/update/urn:li:activity:6817818955353243649/

Part of the reason that I had thought about this so much was due to a counseling class I took that consisted of looking at various types of counseling methods through the question “What is Your Epistemology of Change?”  The class was part of a series of classes in Family Systems counseling that were taught by the best teacher I ever had.  His name was Michael (Mick) Mayhew, and the classes were part of the Family Systems curriculum at St. Cloud State University in Minnesota, US. 

I don’t know if I ever completely wrapped my head around the meaning of “epistemology” in this context.  The word seems to be generally associated with “knowledge” and so I translated this as “how does change occur and how will we know when change has occurred?”  This was a great exercise for anyone interested in promoting change.  What is Your Epistemology of Change?

A year spent analyzing change

After I finished my Master’s Thesis, I was fortunate to get a year of tutoring in the statistics of life course sociology from Ross Macmillan at the University of Minnesota.  The results of that year of analysis are posted at

The-Nonlinear-Dynamics-of-Criminal-Behavior

I apologize for the academic nature of that paper.  For those not well versed in life course criminology, the references might seem obscure.  I will need to rewrite this if I want to reach a broader audience.

How can we assess change when something is always changing?

If you could take one thing away from this paper, it would be the images on Page 30.  In those analyses, I looked at changes in risk scores over three measurement cycles.  The first measurement cycle contained a “get to know you” change due to raters getting to know the offenders better.  The changes in risk scores between assessment 2-3 and assessments 3-4 show that change is constant.

Change is CONSTANT!

This is so important.  People are constantly changing.  This challenges our current scientific paradigms. 

How do you create change in a constantly changing system?  We don’t study the changes that occur because we tend to focus on two point in time measurements.  We almost totally ignore the constant hum of “natural change” in our scientific studies.  

For an exception to this trend, read Fleeson’s 2001 paper.

https://personality-project.org/revelle/syllabi/classreadings/fleeson.2001.pdf

Fleeson shows that personality traits are not stable.  They are constantly shifting over the course of the day.  My work shows that traits are shifting over months as well.

We need to change our research paradigms

We need to stop focusing on single point in time measures.  We need to start looking at how traits change over time.

 

 

Posted by Thomas Arnold