The unreliability of DJN techniques

Conclusion

Without a systematic study of a local community, it is not possible to estimate, with sufficient precision, the percentage of the Jewish community that has either a DJN or a listed telephone number. The consequences of incorrect assumptions concerning either are literally multiplied. DJN techniques all require that the counted directory listing of DJN be multiplied to estimate Jewish persons with non-DJN and to estimate Jewish persons with nonlisted numbers. If the estimates of these values are incorrect, then the counted DJN are multiplied by incorrect numbers, producing wildly unreliable estimates.

To use DJN, therefore, requires that a systematic study be conducted first, but it is to avoid such studies that DJN techniques are even considered. Does this mean that DJN are worthless as a technique for estimating local Jewish populations? Yes, with one possible exception, to judge from the Hartford experience: While both DJN techniques greatly overestimated the Jewish population of every one of the twenty-six towns that constitute the Hartford area, theirrankings of towns were highly congruent with the sample survey’s ranking (Rho = .91). Thus, within metropolitan areas, DJN counts may provide a reasonable basis for ordering subareas relative to one another. More evidence concerning the consistency of DJN rankings of subareas will be most welcome. However, the absolute size of subareas, as indicated by DJN estimates, should be regarded with suspicion.