Module 4 Key Terms and Definitions

dependent variable: the “outcome” variable in intervention research, sometimes referred to as the “y” variable, hypothesized to be influenced by the independent variable.

independent variable: the “input” variable in intervention research, sometimes referred to as the “x” variable, hypothesized to influence the outcomes (dependent variable).

latency: the stretch of time before something occurs, such as an event or onset of change.

nonoverlap of all pair (NAP) analysis: a form of analysis based on all possible pairwise combinations of data points, comparing two phases of a single system design, resulting in a Z-statistic with a p-value that informs the decision to reject or fail to reject null hypothesis of no meaningful change/difference.

nonparametric analysis: statistical analyses not based on the same set of assumptions about the data that parametric analyses require.

omnibus test of significance: a statistical test that measures overall significance of a set of explanatory (independent) variables without distinguishing which one or ones contribute to the solution—additional post hoc analysis are needed to make the distinctions by individual variables.

paired t-test: a parametric statistical analysis where the dependent variable data were collected longitudinally at two points in time (or otherwise nonindependent on the independent variable), using the t-distribution and calculation of a t-statistic based on shared variance.

two-standard deviation band method: an approach to statistical analysis of single system design data comparing the baseline phase band of values two standard deviations above and below the mean to the intervention phase where two or more consecutive data points falling outside the band indicate meaningful change.

x axis: horizontal dimension on a line graph (and some other forms of bivariate graphs)

y axis: vertical dimension on a line graph (and some other forms of bivariate graphs)