The answer to this question lies in the fact that the model (Equation (2)) contains a variable to distinguish between the two surveys used in the modeling (DEMS and MIDAS), which allows one to make either "DEMS" estimates or "MIDAS" estimates.
The researchers conducted six surveys, using the online Mechanical Turk public-opinion tool, between June 2015 and November 2015.
The differences between two surveys were compared using χ test (category variables) and Wilcoxon tests (continuous variables).
To facilitate fair comparison between the two surveys we used the same methodologies that were used in 2008.
The construct validity was tested by comparing indicator level scores between virtual and physical surveys using three statistical measures i.e., % agreement, Kappa Coefficient, Intraclass Correlation Coefficient.
Differences in characteristics between the two surveys were evaluated using Fisher's exact test.
The puzzle is the enormous divergence between the two surveys that are used by the Bureau of Labor Statistics to measure job creation and unemployment.
Today, surveys use seven categories and sometimes even more.
Both surveys used three cuff sizes.
Statistically significant changes in MNCH indicators between the two survey periods were assessed using Pearson's chi-squared statistics, adjusted for survey design.
