Dietary assessment and physical activity measurements toolkit

Glossary of terms

 

  • Accuracy – the extent to which the measurement is close to the true value. A measurement can be reproducible (or precise) but at the same time inaccurate. This will occur if there is systematic bias in the measurements.
  • Bias – a condition that causes a result to depart from the true value in a consistent direction. If bias is present, this reduces the accuracy of a measurement by altering the mean or median value. There are a number of types of bias; the main ones being selection bias and measurement bias.
  • Selection bias - occurs when the subjects studied are not representative of the target population about which conclusions are to be made.
    Possible sources of selection bias include: differential unavailability due to illness, migration or refusal (nonresponse bias). The people who refuse to take part in the survey or drop out may have characteristics that differ from those of the responders.
  • Measurement bias - results from poorly measuring the outcome you are measuring. There are many sources of error which can lead to measurement bias. These include:
    • Social desirability bias: the tendency to provide answers in such a way to avoid criticism e.g. 'good' foods are often over-reported and 'bad' foods under-reported.
    • Interviewer bias: may occur if different interviewers probe for information to varying degrees, record subject responses incorrectly or intentionally omit certain questions.
    • Recall bias: arises with the tendency of subjects to report past events in a manner that is different between the two study groups.
  • Generalisability – describes the extent to which scientific findings can be applied to settings other than that in which they were originally investigated e.g. different population groups, other time periods.
  • Reliability (reproducibility) – refers to the consistency of a measure. A test is considered reliable if similar results are obtained repeatedly. The reproducibility of an assessment method is a function of measurement errors and variability introduced by a variety of other confounding factors (e.g. age, season)
  • Test-retest reliability – a method of assessing the reliability of a questionnaire by administering the same or parallel form of a test repeatedly.
  • Internal consistency reliability – how well items that reflect the same construct yield similar results.
  • Inter-rater reliability – the consistency of measurement obtained when different raters/observers independently administer the same test or use the same method of assessment. This is particularly important in investigator-determined subjective methods e.g. interview or observation.
  • Reproducibility – the degree to which repeated measurements of the same variable give the same value. A closely related term is:
  • Precision - A measure of the closeness of a series of measurements of the same variable
  • Systematic bias – is a bias of a measurement system or assessment method which leads to results which are consistently too high or too low.
  • Random bias - arise from random fluctuations in measurements and can be reduced by repeating the measurement a number of times and averaging the multiple measurements using statistics e.g. median.
  • Validity – the extent to which a measurement instrument reflects what it is intended to measure.
  • Internal validity – concerns whether the study provides an unbiased (valid) estimate of what it claims to estimate. Assessment measures may only have ‘internal’ validity indicating that the results are valid only for the group of participants studied.
  • External validity - refers to the extent to which the results of a study are generalizable to other groups of individuals not involved in the study.
  • Content validity - the degree to which the items on a measurement tool represent the entire range of possible items being measured.
  • Criterion validity – Criterion-related evidence of validity is demonstrated by examining the relationship between a surrogate measure (typically self-reports of physical activity) and a criterion measure e.g. in physical activity research, the criteria that have been typically used include indirect calorimetry, doubly labeled water and direct observation. 
  • Construct validity – the degree to which a test measures the theoretical construct it intends to measure e.g. physical activity
  • Convergent validity - measures of constructs that theoretically should be related to each other are related to each other. Examined by having several different instruments measure the same construct with a high degree of agreement or concordance between instruments.
  • Relative validity – the extent to which measurement instruments of the same kind (e.g. 24-hour recall & food frequency questionnaire) agree
  • Face validity – If a test has face validity then it looks like a valid test to those who use it

 

Web design by Studio 24