Research Commentary---Too Big to Fail: Large Samples and the p-Value Problem
Citations
543 citations
Cites background from "Research Commentary---Too Big to Fa..."
...In addition to statistical significance and co-efficient signs, one may also need to consider effect sizes and variance when testing hypotheses on big data sets (Lin et al., 2013; George et al., 2014)....
[...]
434 citations
401 citations
250 citations
References
1,764 citations
1,747 citations
1,377 citations
"Research Commentary---Too Big to Fa..." refers background or methods in this paper
...Too Big to Fail: Large Samples and the p-Value Problem Mingfeng Lin, Henry C. Lucas Jr, Galit Shmueli To cite this article: Mingfeng Lin, Henry C. Lucas Jr, Galit Shmueli (2013) Research Commentary—...
[...]
...However, to our knowledge, this approach has not been used and there have been no proposed rules of thumb in terms of how such adjustments should be made....
[...]
...…involve a categorical variable can be studied by splitting the data into the separate categories and fitting separate models (Asvanund et al. 2004, Forman et al. 2008, Gefen and Carmel 2008, Ghose 2009, Gordon et al. 2010, Li and Hitt 2008, Mithas and Lucas 2010, Overby and Jap 2009, Yao et al.…...
[...]
...A large sample also enables the researcher to incorporate many control variables into the model without worrying about power loss (Forman et al. 2008, Ghose 2009, Mithas and Lucas 2010), thereby reducing concerns for alternative explanations and strengthen the main arguments if results remain…...
[...]
...Fo r pe rs on al u se o nl y, a ll ri gh ts r es er ve d. conducted robustness/sensitivity analysis, modifying the independent measures (Forman et al. 2008), or the variable structure (Brynjolfsson et al. 2009, Ghose 2009)....
[...]
1,233 citations
1,117 citations
"Research Commentary---Too Big to Fa..." refers background in this paper
...However, when dealing with models like the probit, one has to specify whether an effect size is being calculated Table 2 Interpreting Effect Sizes for Common Regression Models (Vittinghoff et al. 2005) Functional form Effect size interpretation (where is the coefficient) Linear f y = f 4x5 A unit change in x is associated with an average change of units in y . ln4y 5= f 4x5 For a unit increase in x, y increases on average by the percentage 100(e − 1) (û 100 when < 001). y = f 4ln4x55 For a 1% increase in x, y increases on average by ln410015 × 4û /100). ln4y 5= f 4ln4x55 For a 1% increase in x, y increases on average by the percentage 100(e ∗ ln410015 − 1) (û when < 001)....
[...]
...…like the probit, one has to specify whether an effect size is being calculated Table 2 Interpreting Effect Sizes for Common Regression Models (Vittinghoff et al. 2005) Functional form Effect size interpretation (where is the coefficient) Linear f y = f 4x5 A unit change in x is associated…...
[...]