The article “Watchdog Says L.A. Sheriff’s Department Collects Flawed Data, Reports Inaccurate Statistics” appears in the August 8 edition of the Los Angeles Times. In the article, reporter Maya Lau shows how the Los Angeles County Sheriff’s Department misuses statistics. This article shows how biases can guide not only the collection of data and the methodology used to test hypotheses, but also how biases can lead to actual changes in public policy with potentially detrimental or disastrous outcomes.
The Lau article is based on the primary source, the Huntsman report compiled by the Office of Inspector General of Los Angeles. Data was collected using multiple methods, as the report details different raw data measures presumably used and collected by the Sheriff’s Department. In fact, data was collected using different types of software that did not necessarily integrate with one another, according to both Huntsman and Lau. Duplicate entries or multiple events would be counted as one, leading to egregious errors.
The sampling methods used varied depending on the study, often including inmate populations as well as correctional officers. Data collected was quantitative in nature, although the report exposing the inconsistencies and unethical reporting was qualitative, evaluating the...
Combining qualitative methods including observation and survey, the report revealed that the raw data was not collected validly or ethically and was likewise misinterpreted. While there were no errors in how the Lau study was conducted, there were several research errors in how the initial Los Angeles County Sheriff’s Department data was conducted. For example, “the department recorded in an internal report that there were 382 inmate-on-staff assaults in 2015, but the department told the office of inspector general there were 464 of those assaults that year,” (Lau 1).
While Lau does not include charts, tables, graphs, or other visuals to represent the data, undoubtedly the Los Angeles County Sheriff’s Department used visual aids to present their data in impressive ways, in order to inform public policy. In the Huntsman report that Lau bases her article on, ample graphs and charts are used including those showing deputy involved shootings and in-custody deaths. Central tendency issues like mean, median, mode were not overtly discussed in the Lau report, but would have been covered in raw data analyses. Unfortunately, in neither the Huntsman report nor the Lau article do the authors discuss measures of variance or standard…
