This document discusses the importance of close observation of how statistical data is collected and used in organizational settings. It argues that to understand the meaning and significance of statistics, one must observe how the data is measured and assessed in the context of practical decision making, not just scientific methodology. It cites a study examining how members of a university statistical unit do their work as an example. It also summarizes Garfinkel's work observing how staff in a psychiatric clinic recorded applicant information, noting that key data fields like race, occupation and education were only filled in about a third of cases.
This document discusses the importance of close observation of how statistical data is collected and used in organizational settings. It argues that to understand the meaning and significance of statistics, one must observe how the data is measured and assessed in the context of practical decision making, not just scientific methodology. It cites a study examining how members of a university statistical unit do their work as an example. It also summarizes Garfinkel's work observing how staff in a psychiatric clinic recorded applicant information, noting that key data fields like race, occupation and education were only filled in about a third of cases.
This document discusses the importance of close observation of how statistical data is collected and used in organizational settings. It argues that to understand the meaning and significance of statistics, one must observe how the data is measured and assessed in the context of practical decision making, not just scientific methodology. It cites a study examining how members of a university statistical unit do their work as an example. It also summarizes Garfinkel's work observing how staff in a psychiatric clinic recorded applicant information, noting that key data fields like race, occupation and education were only filled in about a third of cases.
close ethnographic observation of the settings in which, and occa-
sions on which, the actual decisions involving the statistical data are made. Only then is it possible to see just how the data are “meas- ured,” how their significance is assigned, in relation to the other practical considerations that inform the decisions in question. That surely includes how their factual status is assessed, not as a matter of professional sociological methodology aimed at establishing the data’s reliability and validity for scientific/theoretical purposes, but as a matter of the practical methodology of organizational person- nel charged with getting things done. To say that their meaning is found in the practical, organizational uses to which they are put is not to signal the shortcomings of the official statistics, but, on the contrary, to locate them in the world in which they belong and have their being. Although not concerned with official crime statistics, the study by Mair et al. (2016) of some aspects of the work of members of a statistical unit at a British university exemplifies what we have in mind. Garfinkel makes the essential point at several places in his Studies in Ethnomethodology (1967). As you read this summary of his work on this matter, imagine it is not about staff in a psychiatric clinic recording applicant or patient contacts, but is about police officers writing up their notes, or filling out the occurrence reports follow- ing some incident, or coding the occurrence reports in terms of the reporting categories required by the CCJS; and then tie that to the uses to which the compiled statistics are put in the halls of govern- ment as generally described above. When examining the records of a psychiatric outpatient clinic for the purposes of a research study to determine the criteria by which applicants were selected for treat- ment, Garfinkel (1967: chapter 6, in collaboration with Egon Bittner) reports that “there were few items in our schedule for which we obtained answers.” For example,
race, occupation, religion, and education [were obtained] in
about a third of the cases . . . Of 47 items that dealt with the his- tory of contacts between applicants and clinic personnel we had returns on 18 items for 90 per cent of our cases; for 20 other items we got information from between 30 per cent to none of the cases. (Garfinkel 1967: 187)