Looking for help with a specific technique? Here's an expanded list from our toolkit.
Surveys and psychometrics
Univariate and bivariate statistical methods
Types of regression and related predictive or explanatory multivariate models
Approaches that can enhance modeling
- Survey and questionnaire design
- Scale development: distilling multiple items into themes and quantifying each person's score
- Testing for survey bias, including straightlining; preferences for consistent or extreme or middle responses; social desirability bias; or common method bias
- Validation
- Multitrait, multimethod analysis
- Reliability analysis, including Cronbach’s alpha and Cohen’s kappa
- Exploratory factor analysis (EFA) and principal component analysis (PCA)
- Multidimensional scaling
- Cluster analysis
- Pricing analysis, including Van Westendorp, Gabor-Granger, monadic, vignette, conjoint, market basket, and behavioral-economics techniques
Univariate and bivariate statistical methods
- Z-tests and T-tests of mean differences
- Correlation; manipulation of correlation coefficients using Fisher’s Z; adjustments for attenuation
- Non-parametric measures of association (e.g., phi, Cramer’s V, Somers’ D, Hoeffding’s D)
- Effect sizes (e.g., Cohen’s d, R-squared, eta-squared, f, pseudo-R-squared alternatives, concordance, odds ratios, risk ratios)
- Chi-square testing
- Analysis of proportions and the differences between them
- Analysis of distributions (e.g., normal, bimodal, poisson, lognormal)
Types of regression and related predictive or explanatory multivariate models
- Predictive modeling, predictive analytics, risk scoring, machine learning
- Multiple linear regression
- Analysis of variance (ANOVA)
- Analysis of covariance (ANCOVA) and analysis of gain scores (change scores)
- Multiple logistic regression (including binary and multinomial)
- Least Absolute Shrinkage and Selection Operator (LASSO) regression
- Survival analysis and regression incorporating censoring
- Instrumental variables and Two-Stage Least Squares (2SLS) analysis
- Time series analysis, including autocorrelation and Autoregressive Integrated Moving Average (ARIMA) models
- Random forest models
Approaches that can enhance modeling
- Power analysis
- Variable selection techniques, both automated and intentional, including sequential and stepwise modeling
- Data transformations
- Modeling of nonlinear effects
- Modeling of interaction effects, including Chi-Square-Based Automatic Interaction Detection (CHAID)
- Model diagnostics, including analysis of outliers, leverage, residuals, linearity, and multicollinearity
- Partial and semipartial correlations
- Imputation of missing data (including Multiple Imputation using Chained Equations, or MICE)
- Receiver Operating Characteristic (ROC) curves and Area Under the Curve (AUC)
- Cross-validation (e.g., k-fold)
- Sensitivity analysis
- Monte Carlo simulation
- A wide variety of methods for creating data graphics -- for exploring data, illuminating findings, and instructing or persuading an audience