For oligonucleotide (Affymetrix) experiments such as the Golden Spike, the nature of the design demands heavy statistical intervention.
Table 3 outlines some of the common variables within the parent studies linked by the xTEND project, and whether the combination of these measures was considered to be ideal (I), less than ideal (LI), requiring statistical intervention (S), and/or where measures of interest were missing from one cohort (M).
Note: I, Ideal circumstances for data combination; LI, Less than ideal circumstances for data combination (e.g., data re-coding required); S, Statistical intervention required for data combination; M, Missing from one sample or measures not comparable; ^ One subscale was missing from AQoL-6D at ARMHS baseline and imputation was required.
Our approximate adjustment for clustering in this trial using the ICC from Rojas 2006 COL suggests this difference may not reach standard levels of statistical significance (intervention 2/3810 (0.05%); control 117/3815 (3.1%); RR 0.02, 95% CI 0.00 to 1.48, one trial, 7625 participants in 12 clusters, Analysis 2.1).
Data collected from each study include sample population, subpopulations (age, gender, race), survey employed, average age or age range of respondents, measure(s) of binge drinking as outcomes, measure(s) of alcohol prices or taxes as interventions, statistical method(s) employed, control variables included in the model (e.g., income, demographics), and robustness tests.
Because of the small effect sizes expected when comparing instruments rather than interventions, statistical power may be marginal at best.
For a scientific efficacy confirmation of drug candidates or other medical interventions, statistical testing is performed to identify significant differences between test groups.
In our view, the timing of the LSCTC intervention was a factor in why TRACS found no difference in clinical or statistical outcomes between intervention and control groups.
A low P value (or a large chi-squared statistic relative to its degrees of freedom) provides evidence of significant statistical heterogeneity of intervention effects [ 11, 12].
The surveys, designed by the U.S. Institute of Environmental Conflict Resolution, feed a database of environmental conflicts that can help make the (statistical) case for intervention in environmental conflict management.
