Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I want to explain to a non-novice doctor the best way to validate consistency of a measure across a likert-scale. The variable describes an average dose across 5 levels of patients (lower to higher). Whats the easiest way to validate that the dose statistically increases from level 1 to level 2 to level3 level 5?

share|improve this question

This sounds like a contrast-type problem! Let's suppose you're assuming a linear increase between level 1 and level 5:

$L=c_{1}\bar{X}_{1} + \cdots + c_{5}\bar{X}_{5}$,

where $c_{j}$ are the coefficients, and $\bar{X}_{i}$ are the group means (the mean at level $i$)--you're computing the weighted group means. For a linear contrast, you might use the coefficients $-2, -1, 0, 1, 2$, as the coefficients must sum to zero. People usually evaluate these while conducting an ANOVA or related regression analysis. There's a general overview of the concept on wikipedia. I don't know what statistical package you're using, but, if you're conducting your analysis in R, you'll want to look at the contrast section in Julian Faraway's Practical Regression and ANOVA using R, which is both a great reference on the mathematics of what is happening, as well as how to actually use them in R.

share|improve this answer
Is there a statistical test that can check for the case of a linear contrast ? I check Faraways' book on how to do it in R and i was not able to find something. – nkorf Nov 17 '12 at 16:36

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.