Pre-analysis:
-
Research proposal: Assist in establishing clear research aims and hypotheses, identify appropriate methods for statistical analyses, determine sample size calculation
-
Advise on database design to facilitate importing for data analysis: data dictionary, variable naming conventions, coding categorical variables and dates, documenting missing data
-
Creating analysis dataset: converting wide to long, long to wide, merging (including using criteria such as “closest” clinic visit, “best” pulmonary function in year, etc.), macros to create new variables
-
Data cleaning: univariable and multivariable summary statistics and figures to identify outliers and data entry errors
Data analysis:
-
Linear regression,hierarchical models, repeated measures analysis, generalized linear models (eg: logistic regression, Poisson regression, etc.)
-
Multivariate data analysis (eg: cluster analysis, factor analysis, etc.)
-
Design and evaluation of measurement scales (validity, reliability, item response theory)
-
Survival analysis
-
Basic statistical methods (ANOVA, t-tests, ChiSquare test, Wilcoxon Rank Sum test, etc.)
Manuscript preparation:
-
Design manuscript quality tables and figures to summarize results
-
Statistical methods and written summary of results
-
Review of results and discussion to ensure accurate interpretation of all results