Effect of the third person effect - 9
and Russian (the latter in order to allow the inclusion of non-Hebrew speaking
immigrants from the former Soviet Republics).
The sample consisted of 502 individuals aged 18 or older from nine different
development towns in the north and south of Israel. The sample was equally
distributed among men and women. In terms of religiosity, the self-reported
distribution was approximately 20% religious, 40% secular and 40% traditional. Just
over half of the respondents in development towns were born in Israel (51.4%), a
quarter (24.9%) in the former Soviet Republics, and the rest were immigrants from
other countries. Most (58%) were of Mizrahi origin (Jews of Asian or North African
origin), 33.7% were of Ashkenazi origin (European origin) and the rest were of mixed
ethnic background. For reasons detailed below, 30 cases were dropped from the
analyses leaving an effective sample size of 472.
Measures.
Residential mobility thoughts. Three Likert-type items were designed to
measure the extent respondents were committed to continue living in their towns.
Respondents were asked to agree or disagree with the statements: “If I had a chance I
would be glad to leave …. “ (RMT1), “In recent years I have considered or I am
currently considering leaving…” (RMT2), “I would like my children to continue
living in …”, (reverse coded, RMT3). Response categories varied between “strongly
disagree” (coded “1”) to “strongly agree” (coded “5”). Residential mobility thoughts
were modeled as a latent factor influencing these three indicators. Exploratory factor
analysis (principal component, oblique) showed the three items loaded on a single
factor explaining 70 percent of the variance. Cronbach’s
α
for the three items was .79.
Perceived town image. What residents of development towns thought about
how others perceived their towns was measured as a latent variable influencing five
Convention