Understanding Society User Support

Dear Sir/Madam,

I am computing test-retest correlations to assess the temporal stability of risk preference using several measures of risk attitude/risky choice using USoc data (and other panels) and I think I may have detected an issue with the risk data from wave 7.

First, I find that the test-retest correlation of items "f_trriska","f_trhlrisk","f_trflrisk" and the equivalent ones in "g_trriska","g_trhlrisk","g_trflrisk" are negative and substantial (~.45). This is not to be expected after a 1 year interval (i'm conducting a meta-analysis using several household panels with similar data and this would be a first). Second, while, as expected, the items in wave 6 are negatively correlated with age and sex (as is known from the literature and other datasets), the items in wave 7 are positively correlated with age and sex. I've checked an this is true for the spss, state, and tab versions of the data.

Second, I am looking for data referring to Domain-Specific Risk-Taking Scale (DOSPERT) Questions that were supposedly collected in waves 6 and 7 (see Table 60 of Galizzi et al; http://eprints.lse.ac.uk/67554/1/Galizzi_temporal_stability.pdf) but was not able to locate this items. Could you perhaps help with this as well?

Any help would be much appreciated!

Best regards,

Rui Mata