Bad Social Science & TF

There's an interesting newspaper account of a sociologist (James Richardson) who did a study of the COG in 1975 posted at xfamily. The study concluded that "the Children of God (often called simply COG) has become one of the most 'democratic and unauthoritarian' groups to spring out of the now-dispersed Jesus movement" and that the COG "definitely believe in the sanctity of marriage."

How did the sociologist reach these bogus conclusions? Why, on the basis of interviews from visits to 30 colonies! Just remember, by 1975, One Wife and MLs about the flirty fishing ministry were in circulation.

This study represents bad science because it doesn't control for two threats to validity: 1) How does the researcher know that the 30 colonies are representative of the whole group? 2) How does the researcher know that his subjects are telling him the truth?

There's no way of knowing how representative 30 colonies may be, because there's no way to verify the denominator (total number of colonies & membership). Therefore, the scientist can only say that for the 50 or 70 or 125 members surveyed, such-and-such conclusions hold true.

The way to check the veracity of the interview information taken from active members in a purposive, self-selected sample is to interview former members who do not have a motivation to maintain a facade. Former members MAY have a motive to paint a very black, misleading picture of the group, but that problem can be managed by asking for documentation and to look for corroborating testimony. Former members could have given the researcher a copy of One Wife as evidence regarding TF's marriage doctrines. The researcher might have interviewed former members who had no knowledge of each other and lived in different locations.

Leaving former members out of their surveys and interviews is one of the biggest methodological errors made by cult apologists. In the world of peer-reviewed science--particularly anthropology and qualitative research methodology--these sociologists of religion look like bozos.

There's no scientific support for the assumption that disaffiliation from a religious group makes a former member a more biased source of information than an active member. But by assuming BOTH active and former members are biased, researchers can compare information from differing points of view to arrive at a cross-validation of social realities. Ah, the rigors of qualitative methodology...yes, it is time-consuming and enormously tedious to cross-validate categorical information. But you know, good science isn't about proving an opinion is true. It's about establishing facts.

The article reporting on the results of bad science by a University of Utah sociologist can be found at:

http://www.xfamily.org/index.php/Mansfield_News_Journal:_Controversial_Children_of_God_Sect_%27Matures%27