Coding and summarization of analysis results
For each of the three broad age categories, analyses were coded by the four achievement domains (BRS, RC, BMS, and MR). This produced 12 separate CHC- ACH domain summary tables (3 age groups X 4 achievement domains) that can be accessed by clicking on links below. The total number of analyses included by achievement domains, ordered by the three age groups (6-8; 9-13; 14- 19 years), were:  BRS (14, 14, 9); RC (11, 13, 8); BMS (12, 12, 10); MR (13, 14, 12).
As reported in Tables 2a-2l, for each analysis a “vote tally” method of CHC IV significance was recorded. The reported general, broad or narrow IVs were recorded as: (a) significantly related to the achievement domain (X), (b) included in the study but were not significantly related to the achievement domain (O), or (c) were not included as an IV in the analyses (blank space). A significant IV relation was designated if it was reported as significant (t-test; F- test; regression weight; SEM effect size) by the studies researchers. The significance findings were tabulated separately for the manifest variable (MV) and latent variable (LV) analyses, as well as across MV and LV studies (MV+LV).  A ratio of number of significant findings (# s) to total number of analyses which included the IV (# t) was recorded. The “Grand #s / # t” was the primary metric analyzed in the current synthesis. For example, across the 7 BRS analyses (at ages 6-8) that included a broad Gs IV, 5 were significant (5/7; see Table 2a). Each Grand #s / # t ratio was converted to a percent of significant analyses value, reflecting the percentage of times a CHC IV was significantly related to the achievement DV in studies in which the IV was included.
The percent of significant analyses for all CH IVs were organized in graphs (see Figures 1-4:  Figures can be accessed by clicking here). To facilitate interpretation of the summary figures, results were plotted only for CHC IVs that were based on at least two analyses and that had percentages greater than or equal to 20%. Results were further classified as per consistency of significance: high(80% or above), medium(50-79%), low(30-49%), or tentative/speculative.  Our operational definition of consistency of significance was constructed logically based on the post-hoc examination of group ‘breaks” or trends in the summary figures. We recognize these operational criteria may be too arbitrary or liberal for some, while others may argue that they are too conservative. Other scholars are free to invoke different criteria and re-interpret the summary coding tables (see endnote 9).