Skip to main content
. 2023 Feb 21;12:e44627. doi: 10.2196/44627

Table 3.

List of para-data indices.

Index Description
Survey completion

  • Time lag from invitation to survey uptake

  • Time of day survey was taken

  • Missed survey

  • Time from uptake to complete the survey

This category of indices measures aspects of survey completion, such as whether a survey is incomplete and the time spent on completing a survey. Emerging evidence suggests that these indices may be predictive of cognitive impairment [31].
Response time

  • Median browser-side response time

  • Deviations from the median browser-side response time

  • Variability in response times across questions

  • Autocorrelation (inertia) of response times across questions

  • Item response theory person-fit statistics response time patterns

  • Pause rate (no response for >20 seconds)

  • Median pause duration

This category of indices measures aspects of response times such as the median response times, variability, and autocorrelations. A growing number of studies suggest that response times and their derived measures may be associated with cognitive impairment [32,33].
Errors and corrective behaviors

  • Proportion of corrected or changed answers

  • Rate of error messages received (eg, invalid entry)

  • Rate of “back” button use

  • Rate of “next” button use without a completed response

This category of indices measures aspects of errors and corrective behaviors in answering web-based surveys.
Mouse and touch efficiency

  • Number of mouse clicks per page (median) [34]

  • Variability of mouse clicks across screens [34]

  • Total pixel count (inefficiency in movement: “wandering around on screen”) [34]

This category of indices measures aspects of mouse clicks and inefficiencies in mouse movement behaviors.
Keystrokes

  • Median time between keystrokes (for text entries) [34]

  • Variability of time between keystrokes (within a text entry) [34]

This category of indices measures temporal rhythms of keystrokes from keyboard entries by respondents for survey questions.