American Sociological Association

Case 39. Linking Data Sets


Janice Spencer has collected data on how adolescents come to the attention of the juvenile justice system and what happens to them after their initial contact with the system. Her survey instrument and observations have focused on adolescents and their families and depends on self-reported data. Because her larger research agenda connects institutional agendas to the "fates" of adolescents labeled as having behavioral problems, she is also interested in how the police and other officials who come into contact with these adolescents respond. In addition to the study focused on kids and their families, she has been doing an evaluation study under contract for the city that follows police and judges and their decision-making on juvenile cases. Both of these projects have been approved by the relevant institutional review boards. After these are collected, Janice realizes that many of the same cases (i.e., the adolescents) were tracked in both studies and sees great potential in linking the two data sets. However, she realizes that her informed consent procedure did not anticipate that possibility and she becomes concerned about whether this is possible under agreements she has secured from both sets of respondents. The officials in the legal system, with whom she has discussed this issue, assure her that they will do whatever they can to make the data on the system available for research.


  1. What kinds of problems are created by the linking of the two data sets?
  2. Given the great potential in linking these two sources of data, what options could she pursue to insure that informed consent is meaningful in this situation?
  3. Since the legal system officials have offered their assistance, should this alleviate her concerns about linking the contract data and the research data?


The essential problem here is that the research subjects were not informed that the investigator would be bringing together different kinds of information. The original consent form, because the research did not anticipate the potential of linking data sets, only asked respondents to agree to an interview and the use of this information in research. This is a gray area. Many current projects link publicly available data sets or data sets obtained for other reasons to new, original data. Some IRBs have chosen to see this as no problem while others have required that the researchers go back to the original respondents to obtain consent to link the data sets. The cautious position would be to assume that the pursuit of "science" (i.e., that rates of participation may drop because of refusal for linkage) does not outweigh the rights of individuals to be informed about how information collected about them will be used. In fact, it is likely that some data sets that can be made available to researchers (e.g., clinic data) were collected under some, even if minimal informed consent agreement. Researchers should be fully aware of the agreements under which all data were collected.

Get Involved


ASA needs you to serve the discipline

Join a Section

Explore ASA's community of specialists in the discipline

Join ASA

Join or renew your membership