Peterson explains failure to replicate


Students and faculty alike gathered in Karp this past Thursday, Jan. 19, to hear Skidmore College Department of Psychological Sciences Assistant Professor Dr. Daniel Peterson discuss his latest paper “A Failure to Replicate.”

Peterson stands in front of an audience of students and staff to discuss his research, results and his experience with experiment replication. Beejit Sarker | Concordiensis
Peterson stands in front of an audience of students and staff to discuss his research, results and his experience with experiment replication. Beejit Sarker | Concordiensis

This talk is the second to have been presented this term as part of the the Union College Psychology Department Speaker Series and Honors Colloquium.

Peterson received his bachelors, masters, and doctorate from the University of North Carolina Chapel Hill in cognitive psychology. He has also published various papers in the field of cognitive psychology, for which he has received numerous awards.

Peterson’s work in cognitive psychology is popular among the group of faculty belonging to the FDI, or Faculty Development Institute. This group of faculty work in various projects to enhance education here at Union, and Peterson’s work directly correlates to their hopes of enhancing and better understanding how students process and remember information in order to better perform.

Peterson’s current work is related to better understanding which processes allow students to better recognize, process, and remember presented information.

Peterson started off his discussion by stating that in the past year, a study was done replicating one hundred well known psychological experiments, of which only fifty were successful. Peterson discussed that in order for an experiment and their results to be noteworthy and applicable, someone else must be able to reproduce the experiment and get similar results. Only then can the results can be supported and further agreed upon.

Peterson’s experiment focused on the concept of how tests improve memory. In this experiment, he worked with the well known generation effect- correlated to semantic memory- versus the testing effect- correlated to episodic memory-. This experiment worked with how item specific processing works in tandem with the testing effect.

Peterson proceeded to discuss how he had based his own study on Union’s own Gilbert R. Livingston Professor Dan Burns’ negative generation effect, which focuses on the ability to generate information and eventually become worse at remembering.

The first experiment was separated into two phases. The first consisted of subjects being introduced with random target words, and were asked to recall them.

The second phase then called for the subjects to retrieve and restudy the words. In this experiment, Peterson found out that subjects who practiced the retrieval did worse than those who focused on the restudy who did better.

The second experiment included a third phase. This phase is a cued recall, in which a randomized recall was introduced. The results show that the subjects did even better after the third phase.

Peterson’s results supported his theory on how to better educate students. As Peterson stated, “This was a great achievement for myself.”

About a year and a half ago, there was a research paper published in which his experiment was replicated. The researchers had failed to obtain the same results. Peterson, feeling “attacked,” decided to send them every aspect of his procedure and process for the experiment to be replicated again. However, they failed to obtain the same results that Peterson obtained.

Upon further examination, Peterson concluded that the subjects that he had used for his experiment were the reason for the difference in the results. Peterson examined and compiled information on all the subjects including test scores. He concluded that the participant pool was the direct cause for the difference.

Peterson concluded the discussion with a brief question-and-answer session from the audience in order for them to better understand the experiment, the results and what they meant for further research.



Leave a Reply