Skip to main content

Table 3 a - e Relative frequency (as %) of medical graduate responses on competencies in statistics and probability that medical schools need to provide

From: Medical graduate views on statistical learning needs for clinical practice: a comprehensive survey

Topic

n

rank

Required competency

Includes practice

Practice only

Theory and practice

Theory only

Neither

Don’t know

a. Data visualization, Data preparation and manipulation and Software used for statistics

Data visualization

  Graphical presentation of data:

274

1

84.3

32.8

51.6

14.2

1.1

0.4

  Cross tabulating frequencies or percentages:

273

23

37.4

15.0

22.3

28.9

16.1

17.6

  Survival analysis:

277

39

20.9

5.4

15.5

65.0

7.2

6.9

  Forest plots:

274

45

15.0

3.6

11.3

51.8

17.5

15.7

  Receiver operating characteristic (ROC) curves:

275

46

13.6

2.6

11.0

45.4

18.3

22.7

  Cluster analysis:

277

51

6.1

1.1

5.1

52.7

26.7

14.4

  Time series analysis:

272

52

5.9

2.2

3.7

36.4

27.6

30.1

Data preparation and manipulation

  Arranging data in spreadsheets for statistical analysis:

278

2

81.7

25.5

56.1

14.4

2.2

1.8

  Working with subsets of the original data - filtering data:

274

27

33.2

13.9

19.3

38.0

15.3

13.5

  Merging similar datasets:

273

42

17.9

6.2

11.7

38.8

27.5

15.8

  Types of response data:

272

49

9.6

2.9

6.6

41.5

17.3

31.6

Software used for statistics

  Using Excel for statistics-tips and warnings

270

7

64.1

26.3

37.8

23.0

9.3

3.7

  Getting to know the fundamentals of a statistical package such as SPSS:

274

11

57.3

27.0

30.4

22.6

14.6

5.5

b. Ratios, rates and proportions, Summarising data and Probability theory

Ratios, rates and proportions

  Sensitivity, specificity and positive and negative predictive values:

272

4

75.0

26.1

48.9

24.3

0

0.7

  Comparing a study cohort with a general population:

274

22

38.3

13.5

24.8

56.6

4.4

0.7

  Statistical risk estimates:

272

38

23.0

5.2

17.8

62.6

6.3

8.1

Summarising data

  Simple descriptive (or summary) statistics:

272

5

69.9

24.6

45.2

21.7

5.5

2.9

  Confidence intervals:

278

6

65.1

20.9

44.2

34.5

0.4

0.0

  Summarising and analysing missing data:

267

43

17.6

4.9

12.7

50.2

20.6

11.6

Probability theory

  Laws of probability:

278

12

48.9

12.2

36.7

46.4

3.2

1.4

  Concepts and rules of probability:

276

25

34.5

10.2

24.4

61.1

2.9

1.5

c. Reporting statistics and Avoiding bad practice in statistics and exploring study design

Reporting statistics

  Presenting the findings and conclusions of statistical hypothesis tests:

276

8

63.4

20.7

42.0

28.6

5.4

3.3

  Statistical significance, statistical power and some facts about p-values:

275

10

58.5

18.9

39.6

39.6

0.7

1.1

  Valid reporting and interpretation of statistical findings:

274

14

46.0

12.4

33.5

48.4

3.6

2.2

  Statistical effect sizes:

274

15

44.9

11.3

33.6

51.5

2.2

1.5

Avoiding bad practice in statistics and exploring study design

  Principles of good study design:

275

16

42.6

14.2

28.4

5.3

1.8

0.4

  Sample size calculations:

272

18

41.2

14.0

27.2

47.1

8.5

3.3

  Randomization:

275

20

39.6

13.1

26.5

59.3

0.7

0.4

  Misuse of statistics: some statistical blunders and phenomena to look out for in published literature:

274

21

39.4

10.6

28.8

52.9

5.1

2.6

  Different types of study design:

275

26

34.2

9.1

25.1

64.7

0.4

0.7

  Statistical aspects of clinical trials:

272

32

29.4

6.3

23.2

65.8

2.9

1.8

  Retrospective power calculations versus examination of confidence intervals:

275

47

12.7

2.5

10.2

48.7

21.5

17.1

  Cross-over trials:

271

48

11.8

2.6

9.2

70.5

10.3

7.4

d. Procedures explicitly requiring statistical hypothesis testing and Assessing agreement, consistency and correlation

Procedures explicitly requiring statistical hypothesis testing

  Tests of normality:

274

17

41.7

11.3

29.9

43.8

8.0

6.9

  Simple linear regression analysis:

278

19

39.6

14.4

25.2

43.9

11.2

5.4

  Hypothesis tests for a single group of continuous data:

269

24

35.3

10.4

24.9

35.4

13.0

16.4

  One-tailed versus two-tailed hypotheses tests:

271

28

32.8

9.6

23.2

48.7

9.2

9.2

  Hypothesis tests for categorical data:

271

30

32.1

8.5

23.6

35.8

14.0

18.1

  Hypothesis tests for comparing two groups of measurement or ordinal data:

270

33

29.3

5.9

23.0

31.5

17.4

21.9

  Analysis of variance (ANOVA):

270

36

27.0

8.5

18.5

45.6

15.2

12.2

  Multiple linear regression analysis:

278

41

18.3

6.8

11.5

58.3

16.2

7.2

  Analysis of covariance (ANCOVA):

269

44

16.7

5.9

10.8

42.4

23.0

17.8

  Tests of homoscedasticity (or, ‘equality’ of variance):

269

50

6.7

1.5

5.2

30.5

30.1

32.7

Assessing agreement, consistency and correlation

  Correlation coefficients –linear and non-linear:

271

34

29.2

8.1

21.0

53.1

10.0

7.7

  Statistical indices for measuring levels of agreement and consistency:

274

37

27.0

7.3

19.7

55.1

9.9

8.0

  Assessing agreement between two methods of measurement:

268

40

18.7

5.6

13.1

48.9

16.4

16.0

e. Allied topics and Critical appraisals and systematic reviews

Allied topics

  Understanding the proper meaning of an audit:

275

3

81.5

29.1

52.4

16.4

1.1

1.1

  Designing survey questions to support valid statistical analyses:

271

13

48.0

15.9

32.1

36.2

10.0

5.9

  Health-related data sources:

272

29

32.7

9.9

22.8

45.2

7.7

14.3

  Representing socioeconomic status:

278

35

29.1

6.5

22.7

58.6

6.1

6.1

Critical appraisals and systematic reviews

  Conducting critical appraisals:

271

9

62.0

21.0

40.2

30.3

4.1

4.4

  Systematic reviews:

267

31

31.5

8.2

23.2

65.5

2.2

0.7

  1. Note. The column header ‘n’ denotes the number of responses for the given topic, while ‘Practice only’, ‘Theory and practice’ and ‘Theory only’ are abbreviations used for the listed response options ‘carry out the procedure or calculate the statistic(s) using appropriate data’, ‘both of the above’ and ‘understand the theory only’, respectively. Correspondingly, column 4 is formed by combining columns 5 and 6. The entries in this column are included in bold for ease of reference when identifying how the corresponding ranks were obtained in column 3. Percentages are row percentages with the denominator in the calculation pertaining to the number of persons who responded for the listed topic, inclusive of those who responded, ‘Don’t know’. Percentages in the main text of this paper are obtained from combining original frequencies for two response options (e.g. those pertaining to columns 6 and 7). They may therefore differ slightly from those obtained by simply adding the corresponding percentages in the table. This is to avoid rounding errors