Figures
8.1aRegression discontinuity data: linear functional form, additive intervention effect
8.1bRegression discontinuity data: Curvilinear functional form, additive intervention effect.
8.1cRegression discontinuity data: linear functional form, additive intervention effect with unequal effect across levels of assignment variable
8.2bTime series data with nonconstant mean, after first differencing transformation
8.3bTime series data with intervention enabled and disabled (A–B–A design)
8.3cTime series data with intervention, plotting relevant and irrelevant outcomes
8.3dTime series data with interventions at different times for different groups (multiple-baseline design)
13.2Panel A provides a schematic overview of the data types that can be collected via smartphones. Panel B presents other data sources that sensing data can be combined with
15.1An example of a dull spreadsheet representing GPS logs (left) and an extensive list of variables that can be extracted from such logs (right)
15.2The accuracy of the facial recognition algorithm when distinguishing between gay and straight individuals based on their facial images
15.4Facebook ads tailored to appeal to extroverts (left) and introverts (right)
15.5Weekly COVID cases versus scented-product reviews mentioning the lack of smell
15.6A tweet visible to X users and the same tweet downloaded from X’s API and stored in R’s data frame
18.1Hypothalamic–pituitary–adrenal axis and psychological states associated with its activation
18.2Hypothalamic–pituitary–gonadal axis with male and female end organs
19.1Schematic of univariate and multivariate brain responses
19.2General approach for representational similarity analysis
20.1Items may be decomposed into shared general variance, shared group variance, unique specific but reliable variance, and error variance (panel A). An item’s reliable variance can be estimated by test–retest correlations of the items. However, general and group are only defined in terms of other items (panel B)
20.3Residual correlations for ten items taken from the Athenstaedt (2003) data
20.4An example of the Schmid–Leiman transformation applied to two correlated factors of anxiety
20.5Hierarchical cluster analysis using the iclust algorithm for the ten anxiety items from the SPI (panel A) and ten gender role behavior items from Athenstaedt (2003) (panel B)
21.1Four prototypic confirmatory factor analysis (CFA) models of twelve item response variables
21.2Scree plots showing the patterns of the first twelve eigenvalues from exploratory factor analyses of item correlation matrices based on male respondents and female respondents and chance parallel pattern
21.3Scatter plots of factor loadings from male and female exploratory factor analyses
22.1Figures display patterns of change in a phenomenon (Y) over time
22.2Effective sample size at varying levels of within-person similarity in repeated measurements
22.4Depiction of predicted values from multilevel models with random intercepts, slopes, or both
23.2Example of decomposition of nonindependence in the APIM using standardized effects
25.1A path diagram describing a simple mediation model where X indirectly affects Y through M
25.2Examples of confounder, collider, and mediator relationships
25.3Example of confounders aligning with each of the four assumptions for identification of indirect effects
25.4An example of a latent variable mediation model where X is observed (e.g., randomly assigned), and M and Y are latent
26.1The signal detection model describes the discrimination between stimuli of two types
26.3A processing-tree representation of the source-monitoring model
26.4Elements of an MPT model include stimulus (behavior) categories
26.6The processing units of the model from Figure 26.5 are depicted as nodes
27.1Flow of citations through the literature search and screening
