Sabbatical Report: Project End

The goal of my past sabbatical was to conduct a project that studied how simple artificial neural networks learned about uncertain environments. This project was completed this past August 10 when I submitted a 240 page monograph for review at Comparative Cognition and Behavior Reviews. The sabbatical not only involved the actual writing this monograph, but also the collection of new results to be reported. This involved a tremendous amount of new research activity. I developed a number of new mathematical proofs about the relationship between simple artificial neural networks and Bayesian probability. I conducted hundreds of network simulations in order to collect data on how such networks behave in uncertain environments when interactions between cues serve as signals of reward probabilities. I also collected data from 200 Introductory Psychology students to measure their behavior in similar environments, and to compare human probability learning to that of my networks. There are striking similarities between network and human performance, and one of the main goals of my monograph is to use this relationship to support the claim that human probability learning can potentially be modeled by very simple artificial neural networks.

While this study of probabilistic artificial neural networks was the primary activity of my sabbatical, I have also been able to develop a new research project (in collaboration with Cor Baerveldt and his students in the Department of Psychology) on the history of the Center for Advanced Study of Theoretical Psychology at the University of Alberta. In particular, during we have explored archival materials and used our findings to explore the development of the Center’s flagship course ‘Seminar in Theoretical Psychology’ as well as the relationship of Center activities to Cold War social science. Both of these projects have led to manuscripts, one that is currently under review at History of the Human Sciences, the other soon to appear in History of Psychology. I also presented a poster at the 50^th annual meeting of Cheiron on the contents of the books in the collection of the Center which are now included in the D.E. Smith Reading Room at the University of Alberta.

Sabbatical Report: Last Day of 2017

With today marking the end of 2017, I thought I would provide an end-of-year update on my current sabbatical. I provided an update at the end of its first month on August 1, and have not provided much information since then. I have been too busy writing sabbatical projects to take the time to add to this blog!

The primary task for my sabbatical is working on a new book. This book explores the probabilistic behavior of simple artificial neural networks with a combination of citations to classic literature in information theory, probability theory, and cybernetics, with the reports of many formal proofs about network behavior, with detailed presentations of many new computer simulation results, and (eventually) with the report of the results of some new experiments on human probability matching.

The book is taking shape. Today – the last day of 2017 – marked the final tweaking of Chapters 4 and 5. So, the current status of the book is very solid drafts of its first five chapters, amounting to 145 pages that hold 73421 words. The next chapter requires me to scrape off the programming rust and write some new neural network code, which hopefully will not take too long. Once the new code is in place I will proceed with a new chapter on including positive feedback in a learning algorithm, which will then lead into a chapter in which I present the results of an experiment with human subjects that I conducted in September. I’ve been at the University of Alberta for 30 years, and this study marked the first time that I used the subject pool on my own!

The development of the book so far has involved collecting data from hundreds of different networks, not to mention reading a lot of new material. Since beginning this project I have consumed 33 books, and have to read several more to write a sensible version of Chapter 6.

The book project is the primary focus of my sabbatical. However, I have been very busy with another project as well. I have had the good luck to collaborate with my colleague Cor Baerveldt and his graduate students on the analysis of archival material related to the Center of Advanced Study in Theoretical Psychology, which existed at the University of Alberta from 1965 until 1990. We have been very busy with this project; we have two articles currently under review at different journals, and I have built a poster that goes with an abstract that I have submitted to Cheiron. With luck I will be able to present some new historical material when Cheiron holds its 50^th annual meeting in June at Akron, Ohio, the home of a huge archive of psychological material.

In short, the sabbatical project is going well, and I am hoping that 2018 is a happy and productive new year – not just for me, but for anyone who has taken the time to read this post.

Happy New Year!

History, Psychology, and Poster Philosophy

In a few days my wife and I head to Angers, France to participate in the meeting of the European Society for the History of the Human Sciences (ESHHS).  She is one of the speakers in a symposium about the recent controversy concerning the identity of Watson’s ‘Little Albert’, while I am presenting a poster that uses Gantt charts to explore the history of the Department of Psychology at the University of Alberta.

I do not participate in many conferences, but these days when I go to a conference I like to present a poster instead of a talk.  The reason: I prefer to have conversations about research; posters facilitate this while talks do not.  But how does one maximize the ability of a poster to initiate conversations?

My answer to this question is to design posters that have as few words as possible – a poster that is almost exclusively a collection of images.  Minimizing the number of words on a poster reduces the likelihood that someone will simply come up to the poster, read it, and move on without talking.  Remove the poster’s words; you force your audience to ask you what the poster is about – which starts a conversation.

My ESHHS contribution, ‘Using Gantt Charts To Explore The History Of A Canadian Psychology Department’, is one that fully reflects my poster philosophy.  It consists of 14 graphs; the only text on the poster is its title and the various graph labels.  Most of the graphs are Gantt charts like the one described in this previous post.  Two of these Gantt charts are gigantic – they provide the timelines for each faculty member, or for each course offering, throughout over a century of the Department’s history – and all of the others are derived from these two major plots.  It is more of an art piece than a typical scientific presentation.  I really like the look of it, I think that it is my most striking poster ever.

If you are interested, an 11 mb version of the whole poster can be viewed as this PDF file.

I like several things about this poster, which has been ‘test driven’ over the last couple of weeks near the main office of my Department.  First, as you get closer and closer to it, you realize that it contains a lot of information.  It really draws you in.  Second, the graphs are unfamiliar – Gantt charts are not typically seen in this field.  As a result, the poster demands questions, such as 'What do these graphs show?'.  On July 7 I will find out how many questions it draws out!

 

 

University of Alberta Cognition: Late – and Early!

Over the last couple of weeks I have visited the University of Alberta Archives to pore over their copies of past Calendars.  As part of a history project that I am presenting at the European Society for the History of the Human Sciences (ESHHS) next month I have compiled the course offering for Psychology from 1909 to 2015.  This list is composed of 5089 separate entries, which might explain why my eyes are tired and my typing fingers are aching.

The point of collecting this data is to illustrate it (and later analyze it) using Gantt charts; a previous project took this approach to illustrate the various faculty members who have belonged to the Department in its existence for more than a century.  The previous project was pretty laborious; this time around I have been able to automate a lot of it using Excel (and VBA) to organize the data to provide to R (and the Plotrix package) for plotting as a Gantt chart.

This approach can be used to provide some interesting insights into Departmental course offerings.  For instance, the figure below provides the Gantt chart of just those courses related to modern cognitivism:

 

 Examining the Gantt chart above indicates that cognitivism arrived at the University of Alberta in the late 1960s.  The first offering, “Topics in Cognition”, appeared in the Calendar for 1968-69.  The Department’s first hiring of an ‘official’ cognitive psychologist was in 1979 when they recruited Alinda Friedman, who has just retired after becoming the longest serving female faculty member in Department history.  Given that cognitivism arose in the mid to late 1950s, it seems that University of Alberta was a pretty late entrant into the cognitivist movement!

Interestingly, though, this story is incomplete.  One of the earlier courses offered by the Department was ‘Legal Psychology (Psychology 56)’, which appeared in the 1922-23 Calendar, and was last offered in 1939-40.  When it first appeared in the Calendar it was described as a course about “normal and abnormal mental processes in relation to problems of judicial procedure”, and explored topics like motivation of crime, the discovery of guilt, mental deficiency and insanity, and individualization of punishment.  This Calendar description was pretty much unchanged from the creation of this course through the 1930-31 Calendar.

However, the Calendar description of Legal Psychology changed markedly in the 1931-32 Calendar, as the image below demonstrates.  The description is split into two parts, with the second one being very similar to the older entries.  However, the new first part is explicitly cognitive in nature: it includes the phrase “cognitive processes”, and focuses on perception, memory, and problems arising in both of these subtopics.

 Two things interest me about this new description of Legal Psychology.  The first is that it demonstrates a very early arrival of cognitivism at the University of Alberta.  This course description is about a quarter of a century earlier than the cognitive revolution!  The second is that I cannot determine any reason for this particular change.  For instance, there were no new faculty members in the Department whose arrival would have led to such a change.

In short, modern cognitivism arose late at the University of Alberta, although the Gantt chart provided above indicates that it is still healthy.  It was preceded, however, by a course in Legal Psychology that was over a quarter of a century ahead of its time.

 

Visualizing The Effects Of Postsecondary Funding Cuts

What long-term effects might result from the recent provincial cuts to postsecondary funding?  No one is in a position to answer that question.  However, a recent paper of mine explores how to visualize a century of a department’s history, and hints at a possible answer by turning to the not too distant past.

The academic year 2009-10 marked the 100^th anniversary of a psychology department at the University of Alberta.  Around this time, the chair of the department asked a small number of people to gather some information and relics to commemorate this event.  My task became generating a roster of the department’s full-time faculty members for each year of its history.  I began this task by examining rosters published online in the University calendar, as well as a couple of published department histories.  This evolved into spending many, many hours at the Law Library, and later at the Book Depository, going through older (never digitized!) calendars.  In the end, I produced an unwieldy spreadsheet that documented this aspect of department history.

I had two problems.  The first was to manipulate the spreadsheet into a form, preferably visual, for easy access and communication.  The second was to convert my long hours of research into something useful (i.e. a publication).  I solved the first problem by using a particular representation, called a Gantt chart that provided an engaging visualization of department history for public display.  I also leaned on my computational vision background to perform some algorithmic processing of this chart to pull additional information out of it.  This solved the second problem, for I wrote a paper about these visual methods, a manuscript that recently appeared in the journal History of Psychology.

Below you will find one of the less challenging analyses of the Gantt chart.  It graphs the total number of faculty members in the department for each year of its history.

This graph reveals four different phases of departmental dynamics.  The first is a long period of slow growth in a small department, ranging from 1909 to 1945.  The second is a long period of steady growth between 1945 and 1988.  The department grew in size by about one faculty member per year over this period.  The third is from 1988 to 2001, and marks a dramatic reduction in the number of faculty members.  The final phase is from 2001 to the present, where there has been some recovery of department size, but not to historical levels.

The sudden downturn in department size, the third phase of the graph, is of interest today.  This downturn reflects changes in provincial funding.  This was a time in which per capita funding for Alberta postsecondary institutions declined by between 36% and 46%.  For instance, the Klein budget of 1993 cut $140 million from postsecondary education.  Premier Klein said of that budget that “you have to hunt where the ducks are”, a line recently echoed by the current minister of Enterprise and Advanced Education.

The graph suggests that after the ducks were hunted, the hunters left, along with a sizable number of faculty members who were not replaced.  This was a long lasting effect: two decades after the Klein cuts, department size had not recovered.  Is a similar downward trend on the horizon as a long-term effect of the recent Redford budget?

Links:

Dawson, M. R. W. (2013). A case study in Gantt charts as historiophoty: A century of psychology at the University of Alberta. History of Psychology, 16, 145-157. pdf version: 315 kb