Md2 Block4 Research

Front 1

Chapter 26
The Pearson correlation coefficient between two variables, x and y:

Back 1

-Is dimensionless.
-Takes the same value when the variables x and y are interchanged.
- Takes the value zero when there is no linear association between the two variables x and y.

Front 2

The estimated Pearson correlation coefficient between systolic blood pressure (mm Hg) and age (years) in a sample of 30 middle-aged women from a given community was r = 0.72 (P < 0.001). Hence r2 = 0.52. Select all of the following statements which you believe to be true.

Back 2

-There is substantial evidence that systolic blood pressure and age in these women are linearly related.

-48% of the variability of systolic blood pressure in these women is unexplained by its linear relationship with age. 48% = 100% - 52%

Front 3

Study of health service needs in Sheffield

In a study carried out to assess whether health services reflected needs in the Sheffield area, 16750 subjects completed a postal questionnaire on the presence of angina symptoms (Payne N and Saul C, BMJ 1997; 314: 257-261). These patients were stratified by electoral ward, so that it was possible to assess the correlation between the prevalence of symptoms and hospital admission rates for coronary revascularisation in each area. The prevalence of symptoms in different wards was positively associated with the Townsend score (Pearson's r=0.79, P<0.001), a measure of the level of social deprivation in each ward which is higher in areas of increased deprivation, as was mortality. However, the correlation between admission rates and Townsend score was less clear (Pearson's r=0.47, P<0.01). Select all of the following statements which you believe to be true.

Back 3

A positive correlation between the presence of symptoms in different electoral wards and the Townsend score implies that in areas of lower social deprivation, a smaller proportion of the population have symptoms of angina.

In order to use Pearson's correlation coefficient, the authors made the assumption that at least one of the two variables was Normally distributed.

The significant correlation between admission rates and Townsend score implies that there is a linear relationship between the admission rates and social deprivation.

Front 4

Fifty subjects with alcoholic cirrhosis underwent an interview in order to assess the reliability and validity of historical variables such as duration of sobriety, duration and quantity of drinking and treatment history on the assessment of an individual's alcohol history (Yates WR et al. Alcohol and Alcoholism 1998; 33: 488-494). In addition, a source close to each subject, usually a spouse, was interviewed by a second interviewer, who was blind to the subject's alcoholism history. Duration of sobriety correlated highly between subject and the alternative source (Spearman's r=0.96, P=0.0001) as did the individual's score on the High-risk Alcoholism Relapse Scale (HRAR, Spearman's r=0.72, P=0.0001). Select all of the following statements which you believe to be true.

Back 4

Spearman's correlation has provided a measure of association between the HRAR scores as assessed by the subject and their collateral source.

Front 5

Chapter 27
A middle-aged man has his systolic blood pressure measured as 169 mm Hg when he attends his doctor's surgery. The doctor takes the patient's blood pressure a month later and finds that its value has dropped to 153 mm Hg. The patient has not changed his lifestyle during the month nor has he received any hypotensive treatment.
Select all of the following statements which you believe to be true. The doctor would have taken this second blood pressure measurement before recommending measures to lower the patient's blood pressure because:

Back 5

The doctor believed that the patient's blood pressure would drop because of the regression to the mean phenomenon.

Front 6

The slope of the linear regression line between an explanatory variable, x, and a dependent variable, y, is:

Back 6

- The same as the gradient of the line.

- The average change in Y for a unit increase in x.


- Often called the regression coefficient.

Front 7

Linear relationships

Back 7

The relationship between age and weight in children aged 5-11 years.

Front 8

Relationship between alcohol concentrations in urine and blood

In a study to assess the relationship between the concentration of alcohol in urine (UAC) and in blood (BAC), measured in mg%, the following simple linear regression equation was derived: BAC = -5.6 + 0.811 x UAC (Kuroda N et al. Am J Fornesic Med Path 1995; 16: 219-222). Select all of the following statements which you believe to be true.

Back 8

-BAC is the dependent variable.

-At a urine concentration of 250 mg%, the predicted BAC is 197.15 mg%.

Front 9

Chapter 28
Assumptions underlying linear regression

Back 9

-The mean of the residuals is equal to zero.

-The residuals are Normally distributed.

Front 10

Cadmium levels in the urine of workers at cadmium pigment plant

When investigating a sample of 556 men working in a cadmium pigment plant, it was found that the cadmium in a man's urine (mg/l) was linearly related to the number of years worked in the plant. The estimated slope of the line obtained by regressing the amount of cadmium against years worked was 0.0009 mg/l per year, with 95% confidence interval (0.0006, 0.0012) mg/l. The intercept was not significantly different from zero. Select all of the following statements which you believe to be true.

Back 10

-Both the upper and lower limits of the confidence interval are positive indicating that the slope is significantly different from zero (P < 0.05).

-A man who had worked in the plant for 10 years would, on average, have increased the amount of cadmium in his urine by 0.009 mg/l.

-A man who had worked in the plant for one year would be expected to have 0.0009 mg/l of cadmium in his urine.

Front 11

Study of prothrombin time and prothrombin activation fragment F1.2

Prothrombin time (expressed as the international normalized ratio [INR]) and prothrombin activation fragment F1.2 (which provides an index of in vivo thrombin generation) were measured in 846 patients with atrial fibrillation participating in a stroke prevention study (Feinberg WM et al. Stroke 1997; 28:1101-1106). F1.2 levels were inversely correlated with INR; the following regression equation was obtained. Linear regression was performed after taking an inverse transformation of INR and the following regression equation was obtained: F1.2 = -0.1 + 2.3 (1/INR), r2=0.37, P<0.0001. Thus, F1.2 levels were inversely correlated with INR. Select all of the following statements which you believe to be true.

Back 11

As the INR value increases, the F1.2 value tends to decrease.