Table 4

Adjusted negative binomial regression model for the impact of statistical reporting and analysis on citation frequency.

Variable

Ratio

Standard error

95% CI Lower Bound

95% CI Upper bound

P-value


Intercept

27.25

1.08

23.24

31.95

Journal

▪AGP

3.11

1.11

2.53

3.82

< 0.001

▪AJP

1.77

1.10

1.48

2.11

< 0.001

▪NJP

0.06

1.21

0.04

0.08

< 0.001

▪Reference BJP

1.00

Research question

▪not stated

0.99

1.10

0.82

1.19

0.888

▪stated

1.00

Primary outcome

▪not stated

0.94

1.10

0.78

1.14

0.554

▪stated

1.00

Sample size

▪not reported

0.98

1.11

0.80

1.20

0.855

▪reported

1.00

Incomplete description of procedures

▪yes

0.83

1.10

0.69

1.00

0.048

▪no

1.00

Inappropriate analysis

▪yes

1.05

1.12

0.85

1.31

0.640

▪no

1.00

Incomplete analysis

▪yes

0.95

1.10

0.78

1.15

0.601

▪no

1.00

Overuse of tests

▪yes

1.026

1.10

0.86

1.23

0.778

▪no

1.00

Dispersion parameter

1.75

1.04

1.62

1.88


Nieminen et al. BMC Medical Research Methodology 2006 6:42   doi:10.1186/1471-2288-6-42

Open Data