Article review: How do you assess the quality of educational research articles?

Imagine this. You are about to conduct an innovative educational project and want to get a research publication out of it. What are considered strong methodological qualities of an educational research study? What can you do to improve your chances for publication?

The authors in this study developed and use an instrument to help measure the methodological quality of quantitative studies in medical education. This instrument, the Medical Education Research Study Quality Instrument (MERSQI), was used to show that scores were predictive of manuscript acceptance into the 2008 Journal of General Internal Medicine (JGIM) special issue on medical education.

What is the MERSQI instrument?
The 10-item MERSQI instrument can be divided into 6 domains, each with a maximum score of 3 points. Numbers within parenthesis denote the number of different items assessed in that domain. The maximum score is 18.

• Study design
• Sampling (2)
  
• Type of data
• Validity of evaluation instrument (3)
  
• Data analysis (2)
  
• Outcomes

The specific scoring criteria are seen in this table below, from an earlier publication on the MERSQI instrument (Reed DA et al, Association between funding and quality of published medical education research. JAMA. 2007;298(9):1002-1009). This earlier publication showed that a high MERSQI score correlated with successful funding of the study.

MERSQI scoring criteria (3 points per domain)
Click to enlarge table image.

So what garners a perfect MERSQI score of 18?

• A randomized controlled trial conducted at over 2 institutions
• A response rate of >75%
• Data assessment was done using an objective measure (not self-assessment by study participant)
• When using an evaluation instrument, internal structure validity, content validity, and criterion validity are reported.
• The data is analyzed appropriately and goes beyond just descriptive analysis.
• The outcome measure goes beyond simple survey measures and focuses on instead patient/health care outcomes.
  

Getting a perfect score of 18 is extremely difficult, as evidenced in this article. In the 100 submitted manuscripts to JGIM, the mean MERSQI score was 9.6 (range 5-15.5). Most manuscripts were single-group cross-sectional studies (54%), conducted at a single institution (78%). Few (36%) reported validity evidence for their evaluation instruments.

The mean total MERSQI score of accepted manuscripts was significantly higher than rejected manuscripts (p=0.003).

• Mean score for ACCEPTED manuscript = 10.7
• Mean score for REJECTED manuscript = 9.0

I have found the MERSQI scoring instrument extremely helpful in helping me design my educational research studies. Methodologic rigor is almost always the Achilles heel of rejected educational research submissions.

Reference
Reed DA, Beckman TJ, Wright SM, et al. Predictive validity evidence for medical education research study quality instrument scores: quality of submissions to JGIM's Medical Education Special Issue. J Gen Intern Med. 2008 Jul;23(7):903-7.