Measures

As part of an occupational history questionnaire, participants were asked to provide information about their longest held main occupation. This included information about occupational title, task specifications, and the number of years in this occupation. Main occupation was then matched with the best fitting category listed in the fourth edition of the U.S. Dictionary of Occupational Titles (DOT; U.S. Department of Labor, 1977). In DOT, more than 12,000 occupations have been evaluated based on observations by job analysts. Occupations are classified based on a 9-digit code (e.g., 354.374-010, nurse), and the three digits in the middle represent occupational complexity with data, people, and things, respectively. Scores were coded so that for each dimension a higher value was indicative of higher complexity (ranges for data 0–6; for people 0–8, for things 0–7). The coding and categorization of worker activities into working with data, people, and things has been used (see e.g., Smart et al., 2014; Boots et al., 2015; Feldberg et al., 2016) and validated in previous studies and is therefore a useful tool for classifying work requirements (Kohn and Schooler, 1983; Peterson et al., 2001). Dimensions used in classification of occupations can be seen in Table 1.

Dimensions used in the rating of occupations into complexity of working with data, people, and things.

The computerized tasks used in the present study were programed in E-Prime 2.0 professional (Schneider et al., 2002). In all tasks, participants were instructed to respond as quickly and accurately as possible.

Three tasks were used to measure inhibition. The Flanker task (Eriksen and Eriksen, 1974) was the first task used as an indicator of inhibitory control. A fixation cross (+) was displayed for 2000 ms followed by five arrows shown at the center of the screen (i.e., < < > < <). Participants were instructed to ignore the direction of the flanker arrows and respond to whether the arrow at the center of was pointing to the right (pressed “M” key) or the left (pressed “X” key). In congruent conditions, the central target points in same direction as the flanker arrows. In incongruent conditions, the central target points in the opposite direction as the flanker arrows. The task started with 10 practice trials followed by 96 (6^*16) test trials. Half of the trials were congruent conditions and half incongruent. If no response was given, the stimuli remained for 2000 ms on the screen. Two measures were used as indicators of performance: (1) The difference in mean response time (RT) between congruent and incongruent trials (the so-called Flanker Effect), and (2) the number of total errors.

The Stroop task (Stroop, 1935; Lu and Proctor, 1995) was the second task used as a measure of inhibition. In this task, a fixation was first displayed for 600 ms followed by a “color” word. That task was to identify the ink of the written color word. In congruent trials, the color of the ink in which the word is written matches the color name (e.g., blue written in blue ink). In incongruent trials, the ink did not match the name of the color (e.g., yellow written in green ink). Participants were given two alternative answers, one on each side of the stimulus, and were instructed to press the “M” key if they thought the alternative on the right was the correct, and the “X” key for the alternative to the left. After a response was made, the next trial started. The task started with six practice trials followed by 96 (2^*48) test trials. Two performance measures were used: (1) The difference in mean RTs between congruent and incongruent trials (the Stroop effect), and (2) total errors.

The Simon task (Simon and Wolf, 1963) was the third task used to measure inhibitory control. This version of the task was programed based on the one described by Bialystok et al. (2004). A fixation cross was displayed for 800 ms, followed by a 250-ms blank interval, and then a green or red square was presented either on the right or left side of the screen. The task was to identify the color of the square and press the “X” key, located on the left side of the keyboard, if the square was red, and press the “M” key, located on the right side of the keyboard, if it was green. In congruent conditions, the square was presented on the same side of the screen as the associated response key on the keyboard (i.e., red square on the left side of the screen). In incongruent trials, the square was on the opposite side as the associated response key (i.e., red square on the right side of the screen). Participants had to respond within 1000 ms. Each trial was separated by a 500 ms response-to-stimulus interval. The task started with 20 practice trials followed by 80 (2^*40) test trials. Half of the trials were incongruent trials and half congruent trials. Similar to the other inhibitory tasks used in this study, two measures were used: (1) The difference in mean RTs between congruent and incongruent trials (the Simon effect), and (2) the number of errors.

Three tasks were used to measure switching. The Number-Letter task (Rogers and Monsell, 1995) was the first task used as a measure of switching ability. This was a modified version in which a pair of one number and one letter (e.g., 7A) was presented in one of the four corners on the computer screen. If the pair appeared in any corner at the top of the screen, the participant had to decide if the number was even or odd by pressing the “X” key for odd and the “M” key for even. If the pair appeared in either of the bottom corners of the screen, they had to decide if the letter was in lower or in upper case by pressing the “X” key for a lower case letter and “M” key for an upper case letter. Three test blocks were performed, each of which was preceded by eight practice trials. In the first block (32 trials), stimuli were presented only in the top corners of the screen, and thus, participants categorized only numbers. In the second block (32 trials), stimuli appeared only in the bottom of the screen, and consequently participants made decisions only on the letters. In the third and final mixed block (128 trials), pairs rotated clockwise around the screen and thus a mental shift was required between classification of numbers and letters. Two measures were used: (1) the difference in average RT between switching trials in the third block (mental shift required) and non-switching trials (no shift required) as measure of processing cost, and (2) the number incorrect responses (total errors).

The Color-Shape task, similar to the version used by Prior and Macwhinney (2010), was used as the second measure of switching function. A fixation cross was presented at the center of the screen for 350 ms. Then a blank screen was shown for 150 ms, followed by a figure. In all trials, the figure was either a blue circle, a blue triangle, a red circle, or a red triangle. The task included several blocks. In the first (36 trials), participants were instructed to identify the color of the figure. If it was blue, participants were told to press the “Z” key with their left middle finger. If it was red, they were instructed to press the “X” key with their left index finger. In the second block (36 trials), participants made shape decisions. If the figure was a triangle, participants were instructed to press the “M” key with their right middle finger, and if it was a circle to press the “N” key with their left index finger. Finally, after 16 practice trials, participants performed three mixed-task blocks (3^*48 trials). In the mixed conditions, a pre-cue was shown for 250 ms, and then the stimulus was presented and remained above the figure until a response was given. Participants made color decisions of the figure if the pre-cue was a rainbow, and shape decisions if the pre-cue was a black circle embedded within a black triangle. Half of the trials were switching conditions and half non-switching conditions. Two measures of switching ability were used: (1) The difference in mean RTs between switching and non-switching trials in the mixed task blocks, and (2) total errors.

The Local-Global task (Navon, 1977), similar to the one used by Miyake et al. (2000), was used as the third measure of switching ability. A fixation cross was shown for 350 ms before a figure appeared on the screen, that was either a cross, a triangle, a square, or a circle. Each “global” figure shown on the screen was in turn built up by smaller “local” figures. The “local” figures could either be consistent or inconsistent with the shape of the “global” figure. If it was blue, participants had to decide shape of the global figure. If it was black, participants decided shape of the local figures. If participants thought that the correct answer was “circle” they pressed the “1” key (i.e., 1 line). For a cross they pressed the “2” key (2 lines), for a triangle the “3” key (3 lines), and for square the “4” key (4 lines). Each trial was separated by a 500 ms response-to-stimulus interval and participants performed 38 practice trials before the test including 98 test trials started. Mental switch was required when switching from categorizing a “local” figure to a “global” figure, and vice versa. Non-switch trials were those when participants continued to do the same categorization as in previous stimuli. The test had an equal amount of switch and non-switch trials. The two switch-cost measures used were (1) the difference in mean RTs between switch trials and non-switch trials and (2) the number of errors.

Three tasks were used to measure updating. The N-back task (Kirchner, 1958) was the first task used as a measure of updating ability. In this task, numbers were displayed on the screen, one number at a time. The task was to determine if the number displayed on the screen was identical (yes/no) to the number presented two steps earlier. If “yes,” participants were instructed to press the “M” key. If “no,” to press the “X” key (e.g., 88 = no, 3 = no, 88 = yes, 52 = no, 3 = no, 52 = yes). Each number was displayed for 2500 ms at the center of the screen followed by a 2000 ms blank interval. After 15 practice trials, the participants performed 40 test trials. The number of errors were used as dependent variable in the analyses.

Matrix monitoring (Salthouse et al., 2003) was the second task used to measure the updating function. Two grids were displayed on the screen, one grid at the top half of the screen, the other on the bottom half of the screen. Each grid had 16 boxes, and in both of them there was one black dot. First the participants memorized the location of the dot in both grips for 2500 ms. Then, both grids disappeared and arrows were displayed for 1200 ms in random order both at the top and the bottom of the screen (arrows were not shown not simultaneously). Each arrow was separated by a 250 ms response-to-stimulus interval. Arrows could point in four different directions; left, up, right, or down. If an arrow was shown at the top of the screen, participants had to visualize how the dot moved one step in upper grid. If an arrow was displayed at bottom of the screen, participants had to visualize movement of one step for the dot in the lower grid. Thus, for both grids participants continuously had to update their memory about the location of the dot. At the end of each trial, participants were shown a grid either at the top or at the bottom of the screen with a black dot in one of the boxes. Participants did not receive any information beforehand on what grid (upper or lower) was going to be shown to them. Participants then had to decide whether the dot in the grid was in the correct place (yes = “m” key/no = “x” key) based on the information from the arrows that had been presented. After two practice trials, participants performed 32 test trials. The number of errors was used as the dependent measure in the analyses.

In the Letter memory task (Morris and Jones, 1990), the third task used as a measure of updating ability, a number of letters was shown serially at the center of the screen. Each letter was shown for 2000 ms and participants did not know beforehand the number of letters that was going to be displayed. After two practice trials, 12 test trials were performed, in which the task was to recall the four letters most recently presented, in the correct order. After each trial, participants wrote down the letters they could recall in a test protocol. The number of correct letters in the correct numerical order was used for scoring of results. For easier interpretation in relation to the other executive tasks, the number of errors were used as dependent measure in the analyses.

Age, sex (female = 0, male = 1), and years of education were used as covariates in the analyses.