Class09_2012 -Chi2, Sensitivity, Specificity - Measurement Research Design And Statistics II Week 09 1 Outline for this week Switch to categorical data

Class09_2012 -Chi2, Sensitivity, Specificity - Measurement...

This preview shows page 1 - 10 out of 158 pages.

1 Measurement, Research Design And Statistics, II Week 09
Image of page 1
2 Outline for this week Switch to categorical data analysis Chi squared tests Categorical measures of association Odds ratios Sensitivity, specificity, predictive value
Image of page 2
3 Categorical Data Analysis Categorical Data, Contingency Tables and the Pearson Chi-Squared Statistic
Image of page 3
4 Conceptual introduction We are obviously spending very little time on this This is due to the fact that most of what psychologists do is use scale outcomes...pseudo-continuous variables that we try to predict in a variety of GLM procedures (t-test, regression, ANOVA...)
Image of page 4
5 Conceptual introduction In the disciplines of our clinical colleagues, there has been a much longer history of interest in categorical, often dichotomous outcomes alive or dead sick or cured presence or absence of a disease Increasingly, psychologists have similar interests who has, and does not have, a particular clinical syndrome (depression, anxiety, dementia) psychological predictors of categorical outcomes (mortality, obesity)
Image of page 5
6 Conceptual introduction Biomedical clinicians often measure the effects of their interventions, and the diagnostic sensitivity of their measures, in categorical terms (sensitivity, specificity, positive predictive value, odds ratios) In those cases where we want to offer behavioral interventions as competitors for pharmacological or medical treatments, or at least convince our clinical colleagues that they are equally important, we need to be able to use the same categorical language
Image of page 6
7 Conceptual introduction We’ll begin by talking about categorical analyses in general, but move on to evaluating specialized categorical analysis indexes that speak to both (a) diagnostic accuracy and (b) treatment efficacy. This is complex stuff, and you likely will not remember it unless you take further coursework in epidemiology or categorical data analysis, or your work causes you to predict a lot of dichotomous outcomes
Image of page 7
8 Conceptual introduction The hope is that the materials from today’s class and next week’s constitute a resource you can draw on as you consider these categorical approaches in your own work, read the clinical literature, or critique work from more biomedical disciplines (including, by the way, nursing, pharmacy, medicine, occupational therapy, physical therapy) Do not try to REMEMBER this stuff. Try to make it that you know where to look it up.
Image of page 8
9 Conceptual introduction Two kinds of chi-square tests the “goodness of fit” test looks at a SINGLE categorical variable, and asks whether the observed frequencies in each category differ from expected frequencies these expected frequencies can come from prior expectation (e.g., there will be far fewer blind individuals in the population than blind), or you may assume that all categories will have equiprobability the “independence” test looks at TWO CROSS- TABULATED categorical variables, and asks whether the values assumed by one of the variables is independent of the values assumed by the other
Image of page 9
Image of page 10

You've reached the end of your free preview.

Want to read all 158 pages?

  • Spring '14
  • Marsiske,Michael
  • Chi-Square Test, Pearson's chi-square test, Fisher's exact test

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

Stuck? We have tutors online 24/7 who can help you get unstuck.
A+ icon
Ask Expert Tutors You can ask You can ask You can ask (will expire )
Answers in as fast as 15 minutes