May 2010

Publishing Nutrition Research: Validity, Reliability, and Diagnostic Test Assessment in Nutrition-Related Research

Philip M Gleason, PhD; Jeffrey Harris, DrPH, RD; Patricia M Sheean, PhD, RD; Carol J Boushey, PhD, MPH, RD; Barbara Bruemmer, PhD, RD


The purpose of this article was to discuss aspects that are critical to quality nutrition-related research, namely the reliability and validity of the test assessments used. The researchers also stress the importance of being clear about what one wants to measure and to match that to the appropriate test assessment.

There are a few different tests that can be preformed to assess the reliability of a measurement; it just depends on the type measurement being preformed. Test-Retest reliability is used to test how reliable a certain measurement is by testing it again and comparing the results. Factors such as the cost of the test and the possible influence from the first test attempt should be taken into account before using this reliability test. Inter-item reliability is used to compare items that are testing the same construct to ensure that the same or similar result is found. Inter-rater reliability is essential when using several different individuals to evaluate the topic (ie. food consumption at several middle schools) to ensure that each individual rater would have similar results to the next. Regardless of the method being used, all have statistical measures that can be used to examine the results of the reliability test.

Validity is used to check if a variable is really testing what its intended to measure and similar to reliability, there are a variety of types. Face or Content validity evaluates if the variable actually measures what the researcher wants it to simply by reviewing it, which is a subjective measure. Absolute validity is the highest standard of validity, where a measurement assesses exactly what it was intended to, and this is incredibly rare. Relative validity, however, compares a test method to a reference that has demonstrated a high degree of validity. The last category of validity tests is internal and external validity. Internal validity measures how valid a measurement is within the sample used for the study, such as if a dietary intervention method was effective for the study participants. External validity then applies this to the population as a whole, which may not be possible for all studies.

The article went on to discuss other factors such as false positives, false negatives, etc., showing the wide spectrum of concerns researchers need to address to ensure their study is truly an effective piece of research. These tests are particularly important when trying to detect the presence or absence of a nutrition- or health-related condition. Reliability and validity allow the researchers to address whether the measurement process produced consistent and accurate information that can be helpful to the field of dietetics.


Continuing Education Article:
Publishing Nutrition Research: Validity, Reliability, and Diagnostic Test Assessment in Nutrition-Related Research. 
Philip M. Gleason, Jeffrey Harris, Patricia M. Sheean, Carol J. Boushey, Barbara Bruemmer
Journal of the American Dietetic Association - March 2010 (Vol. 110, Issue 3, Pages 409-419, DOI: 10.1016/j.jada.2009.11.022)

http://download.journals.elsevierhealth.com/pdfs/journals/0002-8223/PIIS0002822309019567.pdf


Supplemental Article:
An Introduction to Qualitative Research for Food and Nutrition Professionals 
Jeffrey E. Harris, Philip M. Gleason, Patricia M. Sheean, Carol Boushey, Judith A. Beto, Barbara Bruemmer
Journal of the American Dietetic Association - January 2009 (Vol. 109, Issue 1, Pages 80-90, DOI: 10.1016/j.jada.2008.10.018)

http://download.journals.elsevierhealth.com/pdfs/journals/0002-8223/PIIS0002822308018956.pdf