Our lab is interested in improving the valid use of physiological measures of stress and health. To this end, we have conducted a review of the literature on physiological measurement in occupational health psychology. Below are our preliminary recommendations for researchers interested in incorporating physiological measures into their research. This page will be updated with relevant resources as we continue our review.
Please cite this page as:
Robertson, M.M., Lefevre-Levy, R., Haynes, N. J., & Eby, L. T. (2018, April). What you need to know about physiological measurement in OHP research.In Lisa M. Kath & Lisa Baranik (Co-Chairs), What you need to know now: Occupational health psychology updates. Symposium presented at the 2018 Society for Industrial and Organizational Psychology Conference, Washington, DC.
Recommendations for Using Physiological Measures in Occupational Health Psychology Research
Enhancing Construct Clarity 1. Prior to choosing to use a physiological measure, carefully consider the nature of the construct of interest, including:
The definition of the construct.
The time course of the construct (e.g., is the construct relatively stable or does it vary over time? If it varies, how often does it meaningfully vary?).
The predictors and/or outcomes of the construct as dictated by theory.
The extent to which validated non-physiological measures of the construct are available, appropriate, and subject to bias.
Enhancing Validity 2. Choose a measure and study design that best reflects the construct you are intending to operationalize.
Provide evidence of convergent, discriminant, and criterion-related validity.
Provide evidence of reliability.
3. Choose an instrument and protocol that assesses the physiological measure in a valid way. 4. Score data in a way that best reflects the nature of the construct (e.g., consider meaningful cut-scores, risk scores, or latent variables). 5. For deductive research, only include physiological indicators in your study if they have demonstrated validity for measuring the construct of interest, even if the instrument you are using collects additional indicators.
Triangulating Findings Using Multiple Methods 6. If a non-physiological measure of the construct is available and appropriate, include this measure along with the physiological measure. 7. Theorize about potential reasons for convergence and/or non-convergence between different physiological measures or between physiological and non-physiological measures of the same construct. Consider including hypotheses regarding the conditions under which physiological and non-physiological measures should or should not converge. 8. If the findings differ across different measures of the same purported construct, provide theoretical and/or empirical explanations for why this may have been the case in the discussion section.
Considering Contaminants 9. Carefully consider potential contaminants to physiological measures and control for these, preferably through the research design. 10. Measure potential contaminants that cannot be eliminated through the research design and consider controlling for these or excluding data before analysis.
Enhancing Transparency in Reporting 10. Report the characteristics of physiological data, including descriptive statistics, distributions, possible outliers, and reliability. 11. Report the effects of control variables, exclusions, outliers, and data transformations on the findings. 12. Report how measures were scored in detail. If using manufacturer algorithms, provide information what is involved in the calculation. 14. Report the substantive meaning of scores.
Discussing Potential Limitations of Physiological Data 15. Discuss potential threats to internal and external validity associated with the use or analysis of the physiological measure.
