Managing Quantitative Data Collection
The data collection process in Phase 4 of the PAIR Project was rather complex, involving a long telephone interview, six daily diary follow-up calls, and the completion of a packet of mailed questionnaires. Keeping track of where all respondents were in the process involved extensive documentation. This section deals with the various means used to manage the data collection.
Tracking the data collection process
Our primary method of tracking subjects involved the construction of posters, on which appeared each individual in the study and their marital and parental status. As each step in the data collection process was completed, the appropriate column was dated and initialed. The posters were kept in a highly visible area where each member of the team could see the progress of data collection. Similar posters were used for data checking and data entry.
Organizing the data collection effort
Folders were also used for respondents to hold the various pieces of data collected. These folders also contained information that interviewers needed to conduct a call (e.g., first names, telephone numbers). Completion of steps of data collection were also logged on the front of the folders. Calendars were included on the front of the folders to aid in scheduling the follow-up calls. Inside the folders was the Telephone Calling Log on which interviewers documented each attempt at contacting respondents. Notes on the call could also be logged. The back of the folder contained a chart to log data checking of the various information collected.
The goal was to complete data checking as soon after data collection as possible. To accomplish this, one senior team member was responsible for collecting folders the morning after telephone calls. This person would log the completion of data collection of the folder and poster and proceed with data checking. If problems were found, the needed information was elicited or clarified on the next scheduled call or as soon thereafter as possible. This procedure was designed to prevent a backlog of data checking at the end of data collection. Calls not completed were either sent to rescheduling or to the holding area for the next scheduled call.
Another senior team member was responsible for selecting the folders for an evening's calls. This way, interviewers knew to go to a central area to receive their assigned workload for the evening. Since interviewers were scheduled to complete a specific type of interview (i.e., long telephone or daily diary calls) the two types of calls were separated. The interviewers completing long telephone interviews received their folders in one area of the office, while interviewers completing daily diary calls received their folders in another.
Our Telephone Calling Schedule was used to ensure that we had enough interviewers to handle the calling load on any particular evening. On a given night, generally needed three interviewers to handle the daily diary calls and two interviewers to cover the long telephone interviews. If, for some reason, we needed extra interviewers, weekly meetings were used to ensure complete coverage.
The PAIR Project at the University of Texas at Austin
Principal Investigator, Ted L. Huston
Page last modified: 29 January 2002