Abstract
With necessary condition analysis (NCA), a necessity effect is estimated by calculating the amount of empty space in the upper left corner in a plot with a predictor X and an outcome Y. In the present simulation study, calculated necessity effects were found to have a negative association with the skewness of the predictor and a positive association with the skewness of the outcome. Also the standard error of the necessity effect was found to be influenced by the skewness of the predictor and the skewness of the outcome, as well as by sample size, and a way to calculate a confidence interval for the necessity effect is presented. At least some of the findings obtained with NCA are well within the range of what can be expected from the skewness of the predictor and the outcome alone.
Highlights
Dul (2016a) has developed necessary condition analysis (NCA) as a tool to investigate whether a factor X can be considered a necessary condition for another factor Y
The standard error of CE-FDH and CR-FDH was calculated for various combinations of sample size, skewness of X, and skewness of Y, and the natural logarithm of these were highly predictable [R2 = 0.71 and 0.87 for log(SE(CE-FDH)) and log(SE(CR-FDH)), respectively] according to the following formulas: Reading the literature mentioned above, we started to suspect that the degree of necessity, as quantified by NCA, might be influenced by the skewness of the predictor and of the outcome, and that the precision of the necessity effect might be influenced by sample size
According to the present simulation, the probability of getting a result indicating a high degree of necessity, as quantified by NCA, increases with a negatively skewed predictor X and with a positively skewed outcome Y
Summary
Dul (2016a) has developed necessary condition analysis (NCA) as a tool to investigate whether a factor X can be considered a necessary condition for another factor Y. In this analysis one calculates the amount of empty space in the upper-left corner when plotting X and Y against each other (Figure 1). According to Dul (2016a), CE-FDH and CR-FDH values below 0.1 could be characterized as small, values between 0.1 and 0.3 as medium, values between 0.3 and 0.5 as large, and values above 0.5 as very large necessity effects
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
