A Comparison of Frequency, Interval, and Time‐sampling Methods of Data Collection

Abstract

Data representing high, medium, and low response rates in constant and nonconstant patterns were generated by electromechanical equipment to determine whether the same data collected by time-sampling, interval recording, and frequency recording would be represented similarly by each method. Results indicated: (1) that time-sampling provided an extremely inaccurate estimate of responding, and (2) that interval recording accurately represented responding of low and medium rates, but grossly underestimated high-rate responding.

An Empirical Method for Determining an Appropriate Interval Length for Recording Behavior

“…However, even when the problems of interobserver reliability (Kazdin, 1977) are satisfactorily overcome, the results obtained using the different recording procedures are not necessarily comparable and all introduce some degree of distortion (Powell, Martindale, & Kulp, 1975;Powell & Rockinson, 1978;Repp, Roberts, Slack, Repp, & Berkler, 1976). In particular, Powell and Rockinson (1978) have demonstrated that interval recording procedures do not permit the frequency of discrete behaviors to be recorded accurately.…”

“…As Jones et al (1975) state: "This difficulty of interpretation affects the scoring of the behavioral record, as, for example, in computing rates-per-minute, simple frequencies of occurrence, and probabilities of sequential interactions" (p. 55). Repp et al (1976) have also demonstrated that with such procedures the utilized interval length can give rise to differences in the rates at which behaviors are reported to occur and in estimates of the duration of those behaviors.…”

“…A search of recent behavioral studies indicated that no study specified the manner of selection, or appropriateness, of the utilized interval length even though it has been demonstrated that this has an impact on the accuracy of data (Powell et al, 1975;Repp et al, 1976). It is likely that an appropriate interval size, one which does not distort data, will be a complex function of the observation scale and the behavioral repertoire of the subjects (Repp et al, 1976).…”

“…It is likely that an appropriate interval size, one which does not distort data, will be a complex function of the observation scale and the behavioral repertoire of the subjects (Repp et al, 1976). If Systems that permit the recording of behaviors in "real time" eliminate within-interval error .…”

Observation Procedure, Observer Gender, and Behavior Valence as Determinants of Sampling Error in a Behavior Assessment Analogue

“…In general, the more frequent the sampling of the population, the more representative were the data obtained. In a comparison of time-and interval-sampling procedures, Repp, Roberts, Slack, Repp, and Berkler (1976) found that the most representative data were obtained in interval recording, whereas time-sampling procedures produced more inaccurate estimates. However, both these latter studies include .. number of deficiencies which limit generality.…”