Telescoping

Telescoping describes a phenomenon that threatens the validity of self-reported dates, durations, and frequencies of events. Respondents often are asked in surveys to retrospectively report when something occurred, how long something lasted, or how often something happened within a certain time period. For example, health surveys often ask respondents the date of the last time, how often, or how many days they were hospitalized during the last calendar year. Answering this type of question requires the respondent to remember exact dates and temporal sequences and to determine whether an event happened within a certain time period. At this stage of the response process, dates or events can be forgotten entirely or “telescoped” forward or backward. While forgetting describes not remembering an event at all, telescoping focuses on errors made by incorrectly dating events that were recalled.

Survey researchers distinguish between two types of telescoping: forward and backward. Forward telescoping occurs when an event is erroneously remembered as having occurred more recently than it actually did. A backward telescoped event is erroneously remembered as having occurred earlier than its actual date. In general, empirical data show that forward telescoping is more likely to occur than backward telescoping.

Why telescoping occurs is not fully understood. Two main theories have emerged in the literature: the time compression and variance theories. However, [Page 887]these theories explain only parts of the phenomenon. The time compression theory focuses only on explaining forward telescoping, arguing that telescoping occurs because of a subjective distortion of the time line. Time is compressed when respondents perceive that events happened more recently than they actually did or a time period seems shorter than the true length of time. This theory also hypothesizes that forward telescoping decreases as the length of the reference period increases. Empirical findings testing this hypothesis, however, have been mixed. Variance theory uses the uncertainty in one's memory about the time of an event as an explanation for telescoping. The theory argues that uncertainty about the timing of an event increases as the elapsed time from the event to when the question is asked expands, explaining both forward and backward telescoping.

Several methods are used to reduce the amount of telescoping. First, “landmark events” can be used to clearly mark the beginning of the reference period. Landmark events are defined as personal or public events that are meaningful and highly salient to the respondent and can therefore provide a temporal structure of events; for example, something happened before or after a car accident. Personally meaningful events are better encoded in autobiographical memory than public events and therefore appear to be better landmark events. These events limit the reference period, provide a temporal structure to the events, and increase accuracy of reports. Bounded interviews are also used to reduce the incidence of telescoping. Used most frequently in panel surveys, bounded interviews permit the interviewer to remind the respondent of his or her reports in the previous interview or to check for overlaps between the current report of events and previously reported events. This technique improves report accuracy by eliminating forward telescoping of previously reported events. Finally, decomposition of a question into several more specific questions has been used. Research has shown that decomposition improves reporting only if the behavior is irregular and dissimilar. Otherwise, decomposing the question can lead to less accurate reports. Thus, the effectiveness of this technique varies over the population, and the use of this strategy should be carefully assessed.

...