Proportional odds regression is used to predict for ordinal outcomes. There are defaults in SPSS that require recodes of all categorical and ordinal variables to correctly interpret the SPSS output. http://www.scalelive.com/proportional-odds-regression.html

Multinomial logistic regression is used to predict for polychotomous categorical outcomes. http://www.scalelive.com/multinomial-logistic-regression.html

Independent samples t-test is used to compare two independent groups on a continuous outcome after meeting the statistical assumptions of independence of observations, normality, and homogeneity of variance. http://www.scalelive.com/independent-samples-t-test.html

#statistics #research #science #researchengineer #scalellc

Type III errors occur as a result of miscommunication and not conducting a hypothesis-driven study. http://www.scalelive.com/type-iii-error.html

Kruskal-Wallis tests are used when the assumption of homogeneity of variance is violated for ANOVA. http://www.scalelive.com/kruskal-wallis-and-homogeneity-of-variance.html

Definitions for several hundred words used in research and statistics with links to webpages for each. http://www.scalelive.com/research-and-statistics-dictionary.html

Compare three or more groups on an ordinal outcome or adjust for violations of statistical assumptions when using ANOVA. http://www.scalelive.com/kruskal-wallis.html

Correlations measure the magnitude and direction of relationships between two variables. http://www.scalelive.com/correlations.html

Test-retest reliability assesses the stability of survey instrument scores across time. http://www.scalelive.com/test-retest-reliability.html

Statistical Power and Large Sample Sizes

Large sample sizes increase statistical power and increase the flexibility of detecting all kinds of effect sizes.

Statistical Power and Extensive Variance of Outcome

Extensive variance in the outcome decreases statistical power and increases the needed sample size.

Statistical Power and Limited Variance of Outcome

Limited variance in the outcome will increase statistical power and decrease the needed sample size.

Statistical Power and Variance of Outcome

Variance of an outcome has implications on statistical power and the needed sample size.

Statistical Power and Large Effect Sizes

Large effect sizes increase statistical power and decrease the needed sample size.

Statistical Power and Small Effect Sizes

Small effect sizes decrease statistical power and increase the needed sample size.

Statistical Power and Effect Size

Statistical power is greatly impacted by the magnitude and variance of the effect size.

http://www.scalelive.com/retrospective-cohort.html

Retrospective cohort designs can yield measures of risk and longitudinal effects.

Logistic regression is used to predict for dichotomous categorical outcomes. http://www.scalelive.com/logistic-regression.html

Statistical Power and Multivariate Designs

Multivariate designs decrease statistical power and increase the needed sample size.

Statistical Power and Within Subjects Designs

Within-subjects designs increase statistical power and decrease the needed sample size.

Statistical Power and Between Subjects Designs

Between-subjects designs decrease statistical power and increase the needed sample size.

Statistical Power and Research Designs

Research design greatly impacts statistical power in applied research and statistics.

Statistical Power and Continuous Outcomes

Continuous outcomes increase statistical power and decrease the needed sample size.

Statistical Power and Ordinal Outcomes

Ordinal outcomes decrease statistical power and increase the needed sample size.