Why is it that Researchers Can’t Talk Straight?
By Steven M. Cohen
A few months ago, the National Ramah Commission released the results of the study I conducted of former campers. Drawing upon the analysis of over 5,000 Ramah alumni, I cautiously wrote: “We can infer that Camp Ramah has been critical to building a committed and connected core of Conservative and other Jews.” And, “Serving as a Ramah staff member is linked with higher levels of Jewish engagement.”
Truth be told, here’s what I really thought: “From all I know, and not just from this study, I’m thoroughly convinced of Ramah’s effectiveness. Serving as a Ramah staff member, after some time as a camper, produces significant Jewish identity outcomes, above and what we’d expect on the basis of the campers’ strong Jewish family backgrounds. The impact grows with each year of staff service.”
Why didn’t I clearly say what I believe? The rhetorical problem isn’t mine alone, but is generic to social researchers. If we’re asked whether X “causes” Y, we answer with scientific legalese. We’ll talk of “evidence of impact,” or “strong associations between X and Y.” Why is it that researchers can’t speak straight? What stops us from writing with the same clarity that our clients adopt when reporting the results of our studies?
The Ramah study, with all its strengths and limitations, provides ample illustration of why we write with so many qualifications. I arrived at my conclusions in part by comparing Ramah alumni with subsamples from the Pew survey and the 2011 NY Jewish Community Study. From the two surveys, I extracted respondents whose parents were in-married Conservative Jews and who attended Hebrew school or day school. The Ramah alumni substantially surpassed both groups in terms of numerous Jewish outcome indicators – from having Jewish friends and spouses, to feeling that being Jewish is very important, to frequently attending Shabbat services.
But these sizable gaps alone don’t conclusively make the case for Ramah’s effectiveness. One reason is that unlike with truly randomized experiments – the kind that are used to test new drugs – we can’t really be sure that the Ramah experience made the difference. The gaps between alumni and the Pew and NY respondents derive in part from “selection bias,” the unmeasured differences between Ramah parents and seemingly comparable parents, to say nothing of the youngsters agreeing to attend Ramah in the first place.
But we have other reasons for caution, one being, “response bias.” Even with a 21% response rate – high by today’s standards – those more enamored of their Ramah experience may have more often answered the call to complete the survey.
Another lies with “list-coverage:” How representative are National Ramah’s mailing lists with the names of those initially invited? The listed alumni may be more Ramah-connected than those who lost contact over the years.
To be sure, less-than-ideal comparison groups, self-selection, response bias and list bias are not unique to the Ramah study, but are endemic to much research on Jewish educational impact. (The Birthright studies constitute a major exception.)
All that said, it turns out that the Ramah data remain intriguing, if not persuasive. While the number of camper years bears little relation to adult Jewish engagement, the former Ramah staff members (72% of the alumni) exhibit more Jewish engagement than those were only campers. And, the more years as staff, the higher the levels of Jewish engagement, controlling for parents’ denomination and day school attendance. It appears that Ramah camping “works” when it leads to Ramah staffing.
Even with these highly suggestive findings, I – as would other social scientists – still resisted speaking of “impact” or “cause.”
But, in exercising such caution, in sticking so closely to the limited survey evidence, social scientists may be selling themselves – the their lay public – short. After all, on most issues we investigate, the most recent survey is simply that – the most recent in a series of studies that generally point in the same direction. The results of a single survey ought not be the sole basis upon which to make reasoned judgments.
In the Ramah example, we have several relevant cognate studies of related issued, sometimes referred to as “side evidence.” They show that adolescence is a period of identity formation; that value education often shows impact, especially when reinforced by parents and peer group; and that alumni of other Jewish camps score high on Jewish engagement, particularly in the ways emphasized by specific camps. In short, it’s fair to conclude that, “Camp Works.”
Beyond side evidence, the researcher of contemporary Jewry benefits from the soft impressions that arise from living among those we study. Over the years, I’ve met so many Ramah alumni among liturgical leaders in Conservative congregations, or at elite gatherings such as a recently held gala for Mechon Hadar or weddings of Jewish leaders’ children. All those repeated and reinforcing observations should be worth something. While many anecdotes aren’t data, and impressions do mislead, to paraphrase the great social scientist Charles Liebman, z”l: When you have a good hypothesis, you don’t have to go looking for confirmatory evidence; it just rolls in.
We researchers who conduct applied social research for the Jewish policy-making public need to learn to better communicate our findings. Perhaps social scientists should think of becoming bi-lingual. Just as the Halacha (law) is augmented by Aggadah (narrative), so too should researchers produce both cautiously stated scientific analyses in scholarly language, as well as more colloquial policy prose for effective communication.
And until we do so, lay readers are advised to translate cautious academic prose into the clearer and more definitive language when writing for the policy-making reading public.
Steven M. Cohen is Research Professor of Jewish Social Policy at Hebrew Union College-Jewish Institute of Religion. He thanks Isaac Bleaman, Paul Burstein, Samuel Heilman, Harold Himmelfarb, Ilana Horowitz, Helen Kim, Laurence Kotler-Berkowitz, Keren McGinity, Deborah Dash Moore, Daniel Parmer, Riv-Ellen Prell, Sam Shuman, Daniel Staetsky, Matt Williams, other members of the Association for the Social Scientific Study of Jewry and Marion Lev-Cohen for their comments on earlier versions of this article.