Secondary use of data

A topic in research methodology

Usually in empirical studies research questions are answered by collecting data that is especially collected. A major part of research design concerns planning the sample and data collection instruments to construct a data set that can be analysed to answer the research questions.

Extant data sets

However, if there were data already available which is suitable for analysing in the study, then it makes sense to use that data. This requires much less work and use of resources for the researchers (and any sponsor such as a funding body) and avoids inconveniencing people who might be asked to provide access and gift data to researchers. In school-based research, this might mean avoiding disrupting classes.

So, using existing data sets, when they are suitable, can be considered more ethical than generating new data. At least, as long as the data is being used in a way consistent with the consent offered by those from whom the data was collected (or with whom it was constructed).

Read about voluntary informed consent

"There is no absolute requirement that data used in a study should have been specifically collected for that study. However, there are two important caveats that should be considered. From a methodological perspective, any study is based around one or more research questions, or some similar focus, and ideally details of the methodology should be designed accordingly. This includes such matters as the type of data to be collected, the amount of data needed, and the approach to sampling. These steps are all short-circuited when data is re-used for a different purpose. It is important therefore to be confident that a data set is appropriate for use for the new purpose. Re-using a data set in this way probably means some compromise over research design features, but does provide considerable savings of time and effort!

From an ethical perspective, data that are collected in schools are a gift to researchers from teachers and pupils. These individuals are the original 'owners' of their feelings, ideas, beliefs, utterances, etc., and, as suggested above, a researcher must get permission for the use of such material in research. (It is normally the parents who give consent for pupils under the age of 16.) For this to be informed consent, the informants must be told the purposes of the research.

Taber, 2013: p.234-5

Examples

Re-using classroom interaction data

"Harrop and Swinson's (2003) study of the nature of questions used by teachers…used the re-examination of a relatively modest amount of classroom recording (5 hours at each of infant, junior and secondary level) – a series of snapshots from a small sample of classrooms. They explain that their research [was designed to] to examine differences in patterns of questioning between the three levels of schooling"

(Taber, 2013: 51)

"…[In] Harrop and Swinson's (2003) study of the nature of questions used by teachers …the data analysed was not originally collected for the study reported in that paper, but consisted of recordings of lessons made as part of an earlier "investigation into teachers' uses of approval and disapproval in infant, junior and secondary schools" (p. 51).

…If teachers gave Harrop and Swinson specific permission to record their teaching to explore their use of approval or disapproval, then the researchers should have sought renewed permission to then re-analyse the data for a new purpose. However, had the researchers' original request been this specific, the teachers' behaviour might have been modified, so it is more likely that the teachers gave consent to the recording being used for more general purposes, e.g., to 'investigate aspects of teacher's classroom talk' or to 'explore aspects of classroom interaction'. If this were the case, that would have covered Harrop and Swinson's (2003) study as well as their original research with the data."

(Taber, 2013: p.234-5)

(Taber, 2013: 51)

Going back to look at gestures

Flood and colleagues (2015) examined the gestures university students made in video-recordings of interviews.

"We selected a corpus of video-recorded interviews that were originally conducted for the purpose of understanding students' experiences in a newly revised learning activity …The interview questions were predominantly focused on the … classroom activity and were designed to collect responses that might inform its subsequent revision. During the activity, students were asked to map the energy levels of the hydrogen atom…to a staircase of a stadium- style lecture hall and to act out the absorption of a photon and the consequent transition between electronic levels. In addition to the interview questions about this activity, students were asked unrelated questions about the molecular geometry of the molecules methane, ammonia, and PF5 to explore what roles the students' hands and bodies play in explaining three- dimensional phenomena."

Flood, et al., 2015, pp.12-13

So, it seems then part of the purpose of the the original interviews did include seeing how learners used their bodies, but the authors tell readers that

"The students were not explicitly encouraged to gesture at any time during the interviews and were not aware that their gestures were a topic of study.

Our interest in the role that gestures play as interactional resources for responsive teaching while students share chemistry ideas emerged from repeated viewings of the video records."

Flood, et al., 2015, p.13

The interest in gestures, then, arose only after the data was collected.

So, this might raise the question of whether the students had volunteered for the research on a broad enough understanding of its purposes to include having their gestures analysed. It is certainly feasible that this was so. It might also be argued that those volunteering to be interviewed were very unlikely to have objected to this secondary analysis. Ideally, however, a reader might like to have it made explicit that the participants had given voluntary informed consent to provide data for the secondary analysis. (There does not seem to be any information in the paper on whether the research had been subject to institutional ethical review; how the participants were enrolled in the original study; what, if any, information they had been given; and what, if any, informed consent was given prior to data collection.)

---

Sources cited:

• Flood, V. J., Amar, F. G., Nemirovsky, R., Harrer, B. W., Bruce, M. R. M., & Wittmann, M. C. (2015). Paying Attention to Gesture when Students Talk Chemistry: Interactional Resources for Responsive Teaching. Journal of Chemical Education, 92(1), 11-22. doi:10.1021/ed400477b
• Harrop, A., & Swinson, J. (2003). Teachers' Questions in the Infant, Junior and Secondary School. Educational Studies, 29(1), 2003.
• Taber, K. S. (2013). Classroom-based Research and Evidence-based Practice: An introduction (2nd ed.). London: Sage.

---