On the practice of dichotomization of quantitative variables.

Department of Psychology, Ohio State University, Columbus 43210-1222, USA.
Psychological Methods (Impact Factor: 4.45). 04/2002; 7(1):19-40. DOI: 10.1037/1082-989X.7.1.19
Source: PubMed

ABSTRACT The authors examine the practice of dichotomization of quantitative measures, wherein relationships among variables are examined after 1 or more variables have been converted to dichotomous variables by splitting the sample at some point on the scale(s) of measurement. A common form of dichotomization is the median split, where the independent variable is split at the median to form high and low groups, which are then compared with respect to their means on the dependent variable. The consequences of dichotomization for measurement and statistical analyses are illustrated and discussed. The use of dichotomization in practice is described, and justifications that are offered for such usage are examined. The authors present the case that dichotomization is rarely defensible and often will yield misleading results.

1 Bookmark
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The associations between endogenous oxytocin levels and emotion recognition in bipolar disorder Objective: Recent studies in patients with Bipolar Disorder (BD) have revealed problems in emotion recognition, specifically for negative emotions, which have been subsequently related to amygdala activity. Previously, the prosocial neuropeptide oxytocin has been shown to be one hormone that alters emotion perception capacities and modulates amygdala response. Accordingly, the aim of this study was to see if plasma oxytocin levels have specific effects on predicting emotion recognition patterns in BD. Methods: Twenty-eight remitted BD patients were recruited for this study and the Vienna Emotion Recognition Task was given. In addition, blood samples were collected for plasma oxytocin analysis. Results: Strong associations were found between fearful emotions and basal oxytocin levels, which were supported by a stepwise regression analysis. Patients with higher levels of basal oxytocin also exhibited greater recognition of fearful emotions. Conclusions: The relationship between recognition of fearful faces and individual endogenous oxytocin levels may contribute to explaining individual differences in social functioning and amygdala dysfunction in BD.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper examines the degree of similarity in assortative pairing for life history strategy, mate value, delinquency, and sensational interests for same-sex friends compared to opposite-sex romantic partners in persons across a variety of nations that are culturally quite different, in spite of some sharing the same language. Specifically, we sampled participants from: (1) Tucson, Arizona, United States of America; (2) Hermosillo, Sonora, Mexico; (3) San Jose, Costa Rica; and (4) Leicester, England. Due to the structure of the data and the nature of our research questions, we used a relatively new statistical tool called the Continuous Parameter Estimation Model (CPEM; Gorsuch, 2005), in order to estimate and compare assortative pairing coefficients between the four cultures. In all four cross-cultural western cultures replications, the inter-rater reliabilities, perceived assortative pairing coefficients, and actual assortative pairing coefficients of all four traits were generally substantial in magnitude, positive, and statistically significant among pairs of social partners as well as pairs of romantic partners.
    Personality and Individual Differences 02/2015; DOI:10.1016/j.paid.2014.12.060 · 1.86 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The present work is an attempt to approach the modeling and visualization of individual opinion towards extremism in a society. We consider a scenario where two opposite concepts are on a dispute for exclusivity, via the effort of the individuals to convince each other of their own oppinion/decision. The power of convincement of each individual over the community depends on its influence, its communicability and on the enthusiasm (extremism) it defends its side. Here we propose a model taking these three factors into consideration. We analyze the system changes in real time by following the agent's extremism parameter of the society, representing its distribution in a graphical fashion and present realizations of the proposed model for a variety of initial configurations (input parameter sets). The results analysis suggest that the most important factor is influence more then the number of individuals defending a position. Moreover, communicability has small or no relevance. Visualization schemes for following individuals interactions are also included.
    Procedia Computer Science 01/2014; 29:2412-2421. DOI:10.1016/j.procs.2014.05.225

Full-text (2 Sources)

Available from
May 23, 2014