Understanding Society User Support: Issueshttps://iserredex.essex.ac.uk/support/https://iserredex.essex.ac.uk/support/support/favicon.ico?15995719382019-01-16T09:46:41ZUnderstanding Society User Support
Redmine Understanding Society User Support - Support #1130 (Resolved): risk items innovation sample waves...https://iserredex.essex.ac.uk/support/issues/11302019-01-16T09:46:41ZRui Mata
<p>Dear Sir/Madam,</p>
<p>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.</p>
<p>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.</p>
<p>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; <a class="external" href="http://eprints.lse.ac.uk/67554/1/Galizzi_temporal_stability.pdf">http://eprints.lse.ac.uk/67554/1/Galizzi_temporal_stability.pdf</a>) but was not able to locate this items. Could you perhaps help with this as well?</p>
<p>Any help would be much appreciated!</p>
<p>Best regards,</p>
<p>Rui Mata</p>