Wednesday, July 25, 2018

An obsessed Canadian pastor creates some junk science

Paul Dirks
Canadian pastor Paul Dirks
I don't know which is worse: A pretentious crackpot study or a group referring to a pretentious crackpot study.

Using the Internet one can “prove” just about anything regarding human sexuality and legions of people will believe just about anything if it seemingly confirms what they “know.” The tonnage of junk is simply amazing. My rules for separating the junk from science are simple.
Science is represented by research conducted by a qualified investigator who publishes his or her findings to a respected academic journal that subjects articles to rigorous peer review.
Even then I remain skeptical.

First we need to visit Family Policy Alliance, an affiliate of Focus on the Family. Stephanie Curry, Esq. writes: Beware of Target Bathrooms. Just in passing, Ms. Curry, Esq, is not licensed to practice law in Colorado where FPA is located. According to Curry, who is the organization's policy manager:
A new and original study, authored by Woman Means Something, has been released showing that reported sexual offenses have risen significantly since Target changed their bathroom policy to be “gender inclusive”.

It’s no surprise that when men have access to women’s and girls’ bathrooms, sexual incidences will increase.
Really? So let us view The Study. The author of The Study is Paul Dirks. Dirks is a Canadian pastor who is obsessed with objecting to laws in Canada that offer protections on the basis of gender identity. I cannot find Dirks' educational background.

Depending upon which link you use The Study is titled Target Study or Gender-Inclusion Policies and Sexual Violence: A Longitudinal Analysis of Media Reports at Target Stores. And the esteemed outlet for The Study? Peer review? Are you kidding?

The pretensions include an abstract styled like real research:

According to The Study, which is quite verbose, incidents of voyeurism have increased at Target Corporation (NYSE:TGT) stores since they announced their gender accommodation policy. This is based on an analysis of news reports. Dirks is comparing the period of January, 2015 through April 2016 with May, 2016 through August, 2017.

There are a number of problems here:

There is a confusion of causation and correlation. Even if there is an increase Dirks cannot prove that it has any relationship to Target's announced policy.

According to the reports that I have read, voyeurs were wanted by police for similar incidents at other stores which may, or may not have policies similar to Target.

Dirks is comparing the 16 months before the policy change with the 16 months following the policy change. A real scientist would have compared the same calendar periods over successive years to eliminate seasonal influences.

Dirks made no effort to determine (from Target's quarterly reports filed with the SEC) comparable foot traffic. Nor did he consider the actual number of Target stores during the study periods.

The basic premise, that newspaper reports (those that Dirks found) provide a reliable indicator, is nonsensical. There are simply too many variables including the volume of other news, the availability of a reporter to view the police blotter, whether or not the reports are credible, and so on.

Dirks methodology was based on Google searches:
Our online searches incorporated a variety of combinations:
  • store signifiers; "Target", "Target store"
  • forensic terms: "police", "blotter", "arrest", "district attorney", "sentenced"
  • sexual offense language; sexual, upskirt, skirt, voyeur, peeping, exposure, indecent, lewd, camera, hidden, assault, rape, grope, touch, masturbate and their cognates
  • date filters
Aside from the potential for selective observation I did a Google News search (you have to use Google News to employ a date filter):

I obtained about a dozen results. Three repeated the same incident. Snopes had investigated at least two Target incidents. One is totally fake and another is real but misleading.

Were I to design a study I would analyze actual arrests during two periods of time. Target has 1,835 stores at this time so that would be a massive undertaking. There are a number of ways that I could create a representative sample but consider that 80 reported incidents, if each was at a different store, represents just 4% of Target stores. In any event I would be working with local police departments rather than media reports.

According to Dirks:
We hypothesized that due to Target's visibility as a company, and the attention given to their gender inclusion policy, it would serve as a signal to sex offenders, explicitly or otherwise, that their stores would serve as an easier context in which to perpetrate bathroom and change room voyeurism against women.
The vast majority of people have no idea what Dirks is talking about. The attention given to Target's policy is primarily in conservative Christian outlets and from the hate group American Family Association (which is boycotting Target).

Moreover, the assumption that the stores serve as easier targets is nonsensical.

Transgender women do not commit these offenses and men simply do not pretend to be transgender to access facilities. I have not done a study. I do not know the nature of these offenses or where in the store they allegedly occurred. Neither does Dirks. I am not qualified to do a study and neither is Dirks.

Dirks' premise is flawed; his methodology is flawed and he is not a qualified researcher. He does a good job of imitating real investigators but he is not one of them.

Moreover, this is not a longitudinal study as that term is conventionally understood. It means observational over time rather than conducting experiments. For example, in a (convenience) sample of 30 people, how many had a flu shot and how many got the flu with and without the shot. The same group would be evaluated the following year and some conclusions could be drawn by sex, age, race, income and so on.

But suppose we dismiss the skepticism and assume that Dirks' findings are something more than selective observation to affirm an hypothesis. Suppose his finding are accurate.

Dirks' objective (which is shared by Stephanie Curry) is to suggest that shopping at Target is dangerous for women and children. Millions of people shop at Target every year. With 80 incidents (occurring in 4% of Target stores) the probability of being victimized by a voyeur seem to be statistically insignificant (I have never been hit by lightning).

The bottom line is that Dirks is not a qualified investigator and his finding have not been published to a respected academic journal that would subject those findings to peer review. This is junk science. In fact it is not science at all. It is merely a reflection of Dirks' extreme — and irrational — transphobia.

Ms. Currie at Family Policy Alliance did not have the intellectual curiosity to evaluate Dirks' findings. Presumably, she had no desire to subject findings that she liked to any critical thinking. There seems to be a lot of that going around in religious conservative circles. Perhaps the absence of critical thinking is contagious.

Related content:

No comments:

Post a Comment

Please be civil and do NOT link to anti-gay sites!

Note: Only a member of this blog may post a comment.