|
nif:isString
|
-
All procedures were approved by a Research Ethics Board at the Robert B. Zajonc Institute for Social Studies, University of Warsaw. Potential participants in the study were informed about its general aim, duration, outline of the procedure and remuneration. They were also told that participation is voluntary, that they can withdraw from the study at any point without telling the reason, and that all data is anonymous and will only be used for reasons related to this study. All participants signed an informed consent form prior to participation.
The study was conducted among 90 University of Warsaw students (73% female) aged 19–25 (M = 21.63, SD = 1.50). They received a small remuneration for participating in the study (ca. 1–2 euros in local currency, exact amount depending on their decisions in the study).
To measure trust we used the trust game [59]—an economic game, in which two players can increase their wealth through the expression and reciprocity of trust. At the outset of the game both players are endowed with some amount of money. The first player decides how much of their initial endowment they would like to send to the second player, knowing that the whole transfer will be tripled when the other person receives it. Then, the second player returns any fraction of currently possessed money (i.e. the initial endowment enlarged by the received transfer) to the first player. After both players have made their decisions, a round of the trust game ends. The amount of money sent by the first player is the measure of their trust toward the partner, while the amount returned is a measure of the second player’s trustworthiness. The trust game is currently the most common behavioral measure of trust [60,61]. In this study the players’ initial endowments in each round were 10 units of experimental currency. The game consisted of 10 rounds, but the participants didn’t know how many rounds there would be. After each round the players’ money was transferred to their accounts in the game and the next round started with new initial endowments of 10 units of experimental currency per person. After the game ended experimental money from the accounts was exchanged into real currency at a rate of 25:1 and given to the participants as remuneration for their participation in the study. Participants could monitor both players’ account balances at all times. The trust game was administered online using dedicated software TGAME, which allows controlling the behavior of one the players by replacing it with a computer strategy. All participants played the role of first players, while the computer always followed the same predefined scenario of the second player’s behavior: in the first “trust-building” stage of the game (rounds one through three) it returned 50% of the money it had; in the second “trust-violation” stage (round four) it kept all the money and returned nothing to the trustor; in the third “trust-recovery” stage (rounds five through ten) it again returned 50% of possessed money. To decrease the probability of uncovering that the second player is in fact imitated by a computer, in all rounds except the fourth (violation of trust) there was a random variability in a 10% range added to its decisions.
To impose cognitive load we used two distinct manipulations, the effectiveness of both of which was confirmed in previous research. In case of the first cognitive load manipulation—memory load condition—participants were asked to memorize a password-like string of seven characters including letters, numbers and symbols [62–65]. The second manipulation involved exposing participants to a disturbing noise (noise load condition). Noise has a moderate but deleterious effect on performance capacity and cognitive tasks, such as the trust game administered in this study, are known to be among the most vulnerable to noise effects [66]. The noise used in the present study was a mix of two musical tracks recorded simultaneously of which one was playing backwards.
When participants arrived at the lab, they were randomly assigned to one of three experimental conditions: memory load, noise load, or no cognitive load condition. Before the primary task begun we administered the cognitive load manipulations. In the memory load condition participants were presented a string of characters and asked to memorize it until they would be asked to write it down. In the noise cognitive load condition they received headphones that emitted a noise and were asked to keep them on until instructed to take them off. In the no cognitive load condition participants did not receive any secondary tasks. Next, all participants played the trust game. When the trust game was over, participants in memory load condition were instructed to write down what they remembered of the password and in noise load condition they were told to take off the headphones. After the trust game and the cognitive load manipulation we also measured participants’ propensity to trust and distrust (a set of generalized beliefs about human nature that determine an initial attitude when interacting with strangers) and their cognitive capacities. Propensity to trust and distrust was measured using Eisenberger’s Expectational Trust Scale (it consists of two subscales—propensity to trust and propensity to distrust) translated by Rozycka and Wojciszke [67]. Cognitive capacities were measured using an abridged version of the advanced Raven’s Progressive Matrices Test [68]; the abridged version included every third matrix from the second series: nos. 1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34; it had a 10-minute time limit. When the study was over participants were thanked, carefully debriefed and given their remuneration. The study was conducted during the daytime (10am-5pm) to avoid large differences in participants’ levels of fatigue. The procedure took about 20–30 minutes.
|
![[This page as RDF]](http://data.gesis.org/somesci/static/rdf-icon.gif){kind=icon}