110 Chapter 5 to a norm group. There were 9 individuals (35 %) with negative thinking as the most prominent profile, 8 had anxiety sensitivity (31 %), 6 impulsivity (23 %), and only 3 had a sensation seeking personality profile (12 %). 3.2 Autocorrelation structures We now explore the autocorrelation functions of each participant. These analyses do not evaluate network structures, but rather characterize the dynamics across all possible timescales within the time-series that will be the input for idiographic network models in the next steps. Appendix C shows the counts and percentages of autocorrelations that were significant at p < 0.01 across all 26 participants, across all possible timescales (lag-0 to lag-59). Autocorrelation functions further demonstrate associations on both short time-lags (e.g., 11.9 % at lag-1 and 5.7 % at lag-2) and longer time-lags. For example, 3.5 % at lag-7 (exactly one-week) and 2.2 % at lag-14 (exactly two weeks). Most within-person bivariate correlations (38.7 % of all bivariate comparisons) can be found at the same-day timescale (lag-0), supporting the contemporaneous timescale as the most appropriate for estimating the networks. 3.3 Idiographic networks For each participant, an idiographic network model was estimated over the complete 60-day timescale. In Figure 1 we present the contemporaneous networks of two participants as an illustrative example. Participant 1 had five bivariate partial correlations that were significantly non-zero. Worrying about the future was positively associated with nervousness (r = 0.09)1 and restlessness (r = 0.08). Restlessness was also positively associated with nervousness (r = 0.07) and acting without thinking (r = 0.02). The latter showed a relatively strong association with doing things that were later regretted (r = 0.13). Similar to participant 1, the association between doing things that were later regretted and acting without thinking was the strongest association for participant 2 (r = 0.26). Daily levels of happiness were negatively associated with restlessness (r = −0.08) and nervousness (r = −0.05). 1 Please note that partial correlation coefficients have been shrunk due to the LASSO regularization technique, so interpretation differs from that of r in traditional linear regression.
RkJQdWJsaXNoZXIy MTk4NDMw