Table 3. Predictors of nutrition-related research manuscripts receiving neutral and negative quality ratings versus positive ratings compared with government-funded intervention research.1 .
| Likelihood of Quality Rating Versus Positive Quality | ||||
| Neutral | Negative | |||
| Factor | OR | 95% CI of OR | OR | 95% CI of OR |
| Funder | ||||
| Government (reference category) | 1.00 | — | 1.00 | — |
| Industry | 1.38 | 0.98–1.952 | 1.90 | 0.95–3.81 |
| Multiple funders | 1.10 | 0.87–1.38 | 0.98 | 0.56–1.71 |
| University/Hospital | 1.543 | 1.21–1.96 | 1.62 | 0.95–2.76 |
| Non-profit | 1.17 | 0.81–1.68 | 1.14 | 0.53–2.45 |
| Not reported | 1.853 | 1.32–2.59 | 4.973 | 2.76–8.95 |
| Type of research | ||||
| Intervention (reference category) | 1.00 | — | 1.00 | — |
| Observation | 0.92 | 0.78–1.10 | 0.77 | 0.52–1.15 |
| Review | 1.384 | 1.00–1.90 | 5.263 | 3.34–8.28 |
Abbreviations: CI, confidence interval; OR, odds ratio.
1
Statistical significance of the model combining funding and type of research to predict quality ratings (χ2 = 118.99, P<0.001, pseudo R 2 = 0.055).
2
Confidence intervals containing the value of 1.0 do not indicate a statistically significant difference between the response and reference category.37
3
P<0.001, based on the Wald statistic.
4
P<0.05, based on the Wald statistic.