"Unobtrusive" Data Collection
When most people decide they need information for a project, they immediately think in terms of surveys and interview projects. Few consider the wealth of information already collected by the organization. These are "unobtrusive measures" because people are not asked to stop what they are doing and fill out forms or speak to consultants.
Types of Data
Different organizations record different types of information. Some examples which have been used in past research include:
- Absenteeism, lateness, and other "behavior" records
- Turnover, accident, and grievance statistics
- Performance information: productivity, costs, re-work, complaints, etc.
- Past survey information
- Letters and memos
When surveys are taken, people expect action to be taken. People may also change their answers, without realizing it, to conform to cultural norms, to match the experimenter's expectations, or for political reasons. Unobtrusive measures avoid these problems.
Unobtrusive measures can be analyzed, as surveys can be; letters, memos, and other "qualitative" information can be content-analyzed, just as interviews and direct observations are.
Sometimes, unobtrusive measures can save a great deal of time and money, because a point can be made using existing data instead of new data. The information may also be more credible than survey findings.
One of the problems of these measures is that the consultant may not know they exist, and may find it difficult to find and use them. There are some potential legal and ethical issues, especially for internal consultants.
Another problem is that the consultant is at the mercy of the organization's record-keeping skills. The data may not be accurate, reliable, valid, or easily used, even if people within the organization find it to be adequate.
Different organizations have different ways of keeping records. For example, is being absent for six days one or six absences? Is an absence of medical reasons treated the same way as a "frivolous" absence? How about "excused" versus "unexcused" absences? Make sure you know the terms.
Consider the problem of validity. Sometimes, the data are altered for political reasons; definitions may change from year to year, or some supervisors may treat individuals differently.
Remember that using the organization's files may raise concerns with the employees, especially if they are not fully informed of the goals of the project, the confidentiality of the data, and the exact uses to which it will and will not be put.
There is tremendous wealth in unobtrusive measures in some organizations, though, in some cases, it is fool's gold. It is worth checking on this information, though, because it is often not used to full advantage by the organization; and it may be more believable to insiders than survey or interview findings.