A study out this week by a Michigan State University researcher claims that couples who met via online dating sites are three times more likely to divorce than couples who met offline. While a sexy finding, the methodology of the study is deeply flawed.
In the analysis the author uses to substantiate her claims that online dating relationships are less stable than offline relationships, the author only accounts for two demographic variables that can be used to explain relationship stability: race and age. Given the relatively low rates of racial intermarriage in the United States, the author's omission of other demographic variables, such as education, religion, and income, is quite curious. In fact, the omission of education, income, and religion from the analysis is particularly troubling, given that the variables were available in the study's dataset and are considered critical to relationship stability and coupling generally.
The author also could have tried to include other variables, such as political affiliation, that could be used to assess the values of partners. Moreover, the author fails distinguish between heterosexual and homosexual couples, who were oversampled in the "How Couples Meet and Stay Together" study used to conduct this research. Due to different legal standards for marriage across states, the author's failure to distinguish between heterosexual and same sex couples is quite a faux paux.
Although the study is being portrayed as inconsistent with prior research on online dating outcomes, the findings most likely would be dispelled if other variables that are traditionally markers of relationship stability were included. Fear not, Tinder, Match, and E-Harmony, prior, more reputable studies suggest marriages formed online are not on life support. Strong academic research, however, may be.