Questions.

  1. The ‘T-Swift’ protein was quantified in the blood of the two groups. Women were found to have higher levels of the protein ( M = 1.062, SD = 0.339, N = 60) than men ( M = 0.528; SD = 0.382, N = 60). Calculate the raw difference effect size (D) and its variance + SE.

  2. Calculate the Cohen’s d for the example above and its variance and SE. Calculate Hedges’ g and its SE.

  3. Convert the Cohen’s d you calculated in step 3 to a Pearson r and its SE. The formula for SE is not provided on the slides but you can get it from Borenstein et al. (2009) chapter 7 (see here)

  4. In a study you see the following table for and association between hay fever and eczema in 11 year old children.

Hayfever
Yes No Total
Eczema Yes 141 420 561
No 928 13525 14453
Total 1069 13945 15522

What is the probability that a child with eczema will also have hay fever? (proportion and/or %). For children without eczema, what is the probability of having hay fever (proportion and/or %). What is the risk difference? Look up the formula for the SE of risk difference, here.

Calculate the Odds Ratio. The ln(OR) and its SE.

  1. In a paper, you find a reported risk difference of 12% between men and women in ever having taken LSD. The authors did not report the SE or raw data but did provide the 95 confidence interval fo their estimate (8.2% to 15.8%). Work out the SE for this estimate, so you can use it in your meta-analysis (tip: use the Cochrane Handbook, alternatively use google).

  2. Bonus : You are reading a paper on the effects of an Angiotensin-Converting–Enzyme Inhibitor, Ramipril, on Cardiovascular Events (simply put: heart attacks) in High-Risk Patients and come across some interesting stats. From the table below calculate the ‘Number Needed to Treat’. Calculate the 95% CI of NNT.

Cardiovascular Event
Yes No Total
Ramipril Yes 651 3994 4645
No 826 3826 4652