Click on the following links. Please note these will open in a new window.

Journal Article 1: Newman, D. A. (2014). Missing data five practical guidelines. Organizational Research Methods, 17, 372–411.

Abstract: Missing data (a) reside at three missing data levels of analysis (item-, construct-, and person-level), (b) arise from three missing data mechanisms (missing completely at random, missing at random, and missing not at random) that range from completely random to systematic missingness, (c) can engender two missing data problems (biased parameter estimates and inaccurate hypothesis tests/inaccurate standard errors/low power), and (d) mandate a choice from among several missing data treatments (listwise deletion, pairwise deletion, single imputation, maximum likelihood, and multiple imputation). Whereas all missing data treatments are imperfect and are rooted in particular statistical assumptions, some missing data treatments are worse than others, on average (i.e., they lead to more bias in parameter estimates and less accurate hypothesis tests). Social scientists still routinely choose the more biased and error-prone techniques (listwise and pairwise deletion), likely due to poor familiarity with and misconceptions about the less biased/less error-prone techniques (maximum likelihood and multiple imputation). The current user-friendly review provides five easy-to-understand practical guidelines, with the goal of reducing missing data bias and error in the reporting of research results. Syntax is provided for correlation, multiple regression, and structural equation modeling with missing data.

Journal Article 2: Larry Tan, K. L., Lambert, P. S., Turner, K. J., Blum, J., Gayle, V., Jones, S. B., Sinnott, R. O., & Warner, G. (2009). Enabling quantitative data analysis through e-infrastructure. Social Science Computer Review, 27, 539–552.

Abstract: This article discusses how quantitative data analysis in the social sciences can engage with and exploit an e-Infrastructure. We highlight how a number of activities that are central to quantitative data analysis, referred to as ‘‘data management,’’ can benefit from e-Infrastructural support. We conclude by discussing how these issues are relevant to the Data Management through e-Social Science (DAMES) research Node, an ongoing project that aims to develop e-Infrastructural resources for quantitative data analysis in the social sciences.

Journal Article 3: Pryjmachuk, S., & Richards, D. A. (2007). Look before you leap and donsagepub.com/stoken/default+domain/: The need for caution and prudence in quantitative data analysis. Journal of Research in Nursing, 12, 43–54.

Abstract: This paper’s aim is to draw attention to the pitfalls that novice and, sometimes, experienced researchers fall into when undertaking quantitative data analysis in the health and social sciences, and to offer some guidance as to how such pitfalls might be avoided. Many health and social science students are routinely instructed that the procedure for undertaking data analysis in quantitative research is as follows: specify hypotheses; collect data and enter it into a computerised statistical package; run various statistical procedures; examine the computer outputs for p-values that are statistically significant. If significant differences are found, jubilation often exists because statistically significant results are deemed to be a clear indicator that something worthwhile (and publishable) has been discovered. This paper argues that this approach has two major oversights: a failure to explore the raw data prior to analysis and an overdependence on p-values. Both of these oversights are routinely present in much health and social-science research, and both create problems for scientific rigour.

Researchers need to exercise caution (‘look before you leap’) and prudence (‘don’t put all your eggs in one basket’) when undertaking quantitative data analyses. Caution demands that, prior to full data analysis, researchers employ procedures such as data cleaning, data screening and exploratory data analysis. Prudence demands that researchers see p-values for their true worth, which exists only within the context of statistical theory, confidence intervals, effect sizes and the absolute meaning of statistical significance.