Statistical and Graphical Analysis of Exam Raw Data

Research output: Contribution to conferencePosterpeer-review

Abstract

Background: Medical students take many examinations during their first two years, thus analyzing exam performance by Course Directors is critical to make sure students perform as expected on exams, and that the exams were written well. Currently, the exam reports from ExamSoft offer detailed performance raw data. We sought to provide professors at OSU-CHS College of Osteopathic Medicine tools for exploratory analysis coupled with graphical visualization for a more informative evaluation of exam performance data.

Methods & Results: Data was taken from a previous exam administered at OSU-COM and uploaded to the R Studio program to generate code for analysis. The code created allows for statistical analysis of the exam, further broken down by each exam writer, allowing for an in-depth analysis. Furthermore, a tool was created allowing for a certain Percent Correct criterion to be set at the discretion of the course director to extract questions below that criterion. This eliminates human error when viewing these questions, and easily marks questions that may need revision.

Conclusion: With these tools, professors will be able to analyze their exams in more depth, allowing for continual improvement of exams administered at Oklahoma State College of Osteopathic Medicine.
Original languageAmerican English
Pages90
StatePublished - 18 Feb 2022
EventOklahoma State University Center for Health Sciences Research Week 2022 : Poster Presentation - Oklahoma State University Center for Health Sciences, Tulsa, United States
Duration: 14 Feb 202218 Feb 2022

Conference

ConferenceOklahoma State University Center for Health Sciences Research Week 2022
Country/TerritoryUnited States
CityTulsa
Period14/02/2218/02/22

Keywords

  • Exams
  • Examination analysis
  • R Studio

Fingerprint

Dive into the research topics of 'Statistical and Graphical Analysis of Exam Raw Data'. Together they form a unique fingerprint.

Cite this