Choosing and interpreting statistics for studies common in primary care
| I want to | Statistical choice | Independent variable | Dependent variable | How to interpret |
| Examine trends or distributions. | Descriptive statistics | Categorical or continuous | Categorical or continuous | Report the statistic as is to describe the data set. |
| Compare group means. | t tests | Categorical with two levels (ie, two groups) | Continuous | Examine the t statistic and significance level. If significant, clearly report which group mean is higher, along with the effect size. |
| Compare group means. | Analysis of variance | Categorical with two or more levels (ie, two or more groups) | Continuous | Examine the F statistic and significance level. If significant, clearly report which group means are significantly different and how (eg, which are higher), along with the effect size. |
| Examine whether variables are associated. | Correlation | Continuous | Continuous | Examine the r statistic and significance level. If significant, describe whether a positive or negative correlation and its strength. |
| Gain a detailed understanding of the association of variables and use one or more variables to predict another. | Regression | Continuous or categorical, may have more than one independent variable in multiple regression | Continuous | Examine the F statistic and significance level. If significant, examine the R square for how much variance the model accounts for. Determine whether each regression coefficient is significant; if significant, discuss the coefficients. |