Table 1
Description of the executive function tasks. This table is based on Table S1 in Hartung et al., 2020.
| TASK DESCRIPTION | DEPENDENT VARIABLE | |
|---|---|---|
| Inhibition | ||
| Animal Stroop (Wright et al., 2003) | Participants were asked to verbally identify animals from drawings based on their body. In the congruent condition, the face of the animal matched the body; in the incongruent condition, the face did not match the body; in the neutral condition, the face area was blank. | Mean reaction time cost for incongruent conditions relative to congruent and neutral conditions |
| Stop signal auditory (Verbruggen & Logan, 2008) | Participants were asked to indicate which way an arrow was pointing, but to inhibit their response if a tone (stop signal) sounded after the presentation of the arrow. | Stop signal reaction time, i.e., subtraction of the average stop signal delay (time between arrow and stop signal presentation) from the mean go reaction time |
| Mickey(Lee et al., 2013) | Participants were asked to indicate on which side of a computer screen a cartoon Mickey Mouse face appeared while ignoring any squares that flashed on-screen before. In the congruent condition, a square flashed on the same side of the Mickey; in the incongruent condition, a square flashed on the opposite side; in the neutral condition, squares flashed on both sides. | Mean reaction time cost for incongruent trials relative to congruent and neutral trials |
| Cognitive flexibility | ||
| Trail making (Salthouse, 2011) | Participants were asked to connect circles containing numbers in numerical sequence and circles containing letters in alphabetical order. In the two simple conditions, only numbers or letters were presented. In the two alternating conditions, both numbers and letters were presented, and the participants had to connect the circles in an alternating sequence (numbers–letters: 1-A-2-B, etc. or letters–numbers: A-1-B-2, etc.). | Switch cost, i.e., the mean reaction time for alternating conditions relative to simple conditions |
| Plus-Minus(Miyake et al., 2000) | Participants were asked to complete simple addition and subtraction problems on paper. In the adding condition, participants had to add 1 to each provided number; in the subtracting condition, they had to subtract 1; in the alternating condition, they had to alternate between adding 1 and subtracting 1. | Switch cost, i.e., the mean reaction time for alternating conditions relative to simple conditions |
| Local Global (Miyake et al., 2000) | Participants had to identify letters and shapes composed of smaller letters and shapes. In the two local conditions, participants had to name the small constituent letters or shapes; in the two global conditions, they had to name the large overall letter or shape; in the alternating condition, they had to alternate between naming local and global letters or shapes. | Switch cost, i.e., the mean reaction time for alternating conditions relative to simple conditions |
| Updating | ||
| 2-Back(Jaeggi et al., 2010) | Participants were asked to watch a sequence of shapes and to indicate when the current shape matched the shape from two trials prior. | Total number of hits (correct responses) minus false alarms (responses incorrectly identified as matching) |
| Running memory (Broadway & Engle, 2010) | Participants were asked to watch a sequence of single letters and to recall the last n letter. | Total number of letters correctly recalled |
| Keeping track (Miyake et al., 2000) | Participants were asked to listen to words falling under four categories and to recall the most recent word from a given category. | Total number of words correctly recalled |
Figure 1
Overview of the latent network model at different focal age points. A = Stroop; B = Stop signal; C = Mickey; D = Trail making; E = Plus-minus; F = Local global; G = 2-back; H = Running memory; I = Keeping track. Green edges indicate positive weights, red edges indicate negative weights, and the thickness indicates the magnitude of these weights.
Figure 2
Fit indices of the 7 latent models. (A) = BIC (Bayesian information criterion) for each model by focal age point. (B) = CFI (comparative fit index) for each model by focal age point. (C) = RMSEA (root mean square error of approximation) for each model by focal age point. Dot line corresponds to threshold for acceptable fit (CFI > 0.95, RMSE < 0.08) and dashed line corresponds to threshold for good fit (CFI > 0.97, RMSE < 0.05; Schermelleh-Engel et al., 2003).
Figure 3
Loadings of the 9 EF variables with age in the unidimensional models.
Figure 4
Overview of the unidimensional latent model at different focal age points. A = Stroop; B = Stop signal; C = Mickey; D = Trail making; E = Plus-minus; F = Local global; G = 2-back; H = Running memory; I = Keeping track. Green edges indicate positive loadings, red edges indicate negative loadings, and the thickness indicates the magnitude of these loadings.
Figure 5
Overview of the latent network model at different focal age points. A = Stroop; B = Stop signal; C = Mickey; D = Trail making; E = Plus-minus; F = Local global; G = 2-back; H = Running memory; I = Keeping track. 1 = Inhibition; 2 = Cognitive flexibility; 3 = Updating. Green edges indicate positive weights, red edges indicate negative weights, and the thickness indicates the magnitude of these weights.