Course site for "Experimental Psychology Lab" 2019
Below you find some proposals for possible replication studies (you can find the corresponding articles on StudIP). You are of course also very welcome to come up with own studies that you would like to replicate. Insofar you are not bounded to these proposals.
Please note: You will encounter the following “situation”:
In most studies several experiments were conducted
Often the design is more complicated than a 2x2 design
The used stimuli are seldom available
The statistical methods differ from what you have learned so far in the course
That’s why you should select only one experiment and try to understand the overall idea of this particular experiment: What are the hypotheses of the author(s)? Which factors play a role? How did they try to test their hypotheses? Once you have understood the general aim you often can break down the hypotheses and for example test only a subset instead of all factors.
Guidelines from designing an experiment until writing a report can be found in “Method_Process&Report”.
Keywords: Anchoring, Numerosity and Number Encoding, Ease of Computation
Hypothesis: The effect of order on perceptions will be moderated by package size and calculation difficulty.
Higher unit price calculation difficulty:
When package sizes are larger, price-item (vs. item-price) presentation order will lead to more negative evaluations of the product offering (e.g., higher unit price judgments, lower trying likelihoods, lower value perceptions,) and to lower choice proportions.
When package sizes are smaller, order effects will not persist.
Lower unit price calculation difficulty:
Experiment 1: Method: The scenario indicated that a company introduced a web-based television service in which payment for access was by the hour. Consumers could purchase as many hours as they wished. We manipulated order by presenting price first (e.g., $300 for 600 hours) or item quantity first (e.g., 600 hours for $300). Larger packages used larger numerosities (600 or 580 hours) relative to the smaller packages (60 or 58 hours). Calculation difficulty was also manipulated between subjects. In the higher (lower) difficulty conditions, 580 hours cost $285.90 or 58 hours cost $28.59 (600 hours cost $300 or 60 hours cost $30). This study used a 2 (order: price-item vs. item-price) x 2 (package size: larger vs. smaller) x 2 (calculation difficulty: higher vs. lower) full factorial between-subjects design. We asked participants to indicate how likely they were to try this offer (1 = not likely at all, 7 = very likely). We also asked participants to indicate how good a deal and how good a value they thought the service provided (both scales: 1 = not good at all, 7 = very good), and used these to create a composite perceived value score. Additionally, participants indicated how many hours of programming were advertised in the offer (text entry), how much it cost (text entry), and whether it was difficult or easy to calculate the price per hour (1 = very difficult, 7 = very easy). These served as manipulation checks for the package size and calculation difficulty manipulations, respectively.
Keywords: Anchoring, Self-construal, Analytic vs. holistic thinking styles
Hypothesis: We predicted that people who define the self more interdependently and who have a more holistic thinking style would be more susceptible to anchoring effects, because, given that they focus more on context and see more connections among different elements, they may be more likely to consider the initial anchor values in relation to their subsequent judgments. In contrast, we predicted that people who define the self more independently and who have a more analytic thinking style would be less susceptible to anchoring effects because they would be less likely to connect anchor values to their subsequent judgments
Method: To measure self-construal, we used Singelis’ (1994) Independent and Interdependent Self-Construal Scales (12 items). To measure individual differences in thinking styles, we used Choi et al.’s (2007) 24-item Analysis-Holism Scale Participants completed 11 anchoring tasks. Each task comprised a comparative question (e.g., “Do you think the length of the Mississippi River is more or less than 200 miles?”) followed by a second question asking participants to make an absolute estimate (e.g., “What do you think the length of the Mississippi River [in miles] is?”). The number in the first question served as the anchor for each task.
Keywords: Desired attitudes, Construal level theory, Motivation, Abstraction, Behavioral intentions
Hypothesis: In Construal level theory, abstract construals are typically seen to be gist-based mental representations focused on the central properties of an object - representations containing lasting, stable, decontextualized features. In contrast, concrete construals are more detailed, including incidental, context-dependent properties. If desired attitudes are represented abstractly, they should be less likely than actual attitudes to be constrained by temporal or situational influences. Thus, we expected higher perceived stability in desired attitudes (versus actual attitudes).
Study 1a: Method: Participants completed self-report measures of their actual and desired attitudes towards a specific topic. On the same page, participants reported their perception of the situational and temporal stability of each type of attitude on that topic. This procedure was repeated for each of four topics. Topics were presented in a fixed order (eating vegetables daily, voting, impulse purchasing, myself). Instruction for participants: “Sometimes the attitudes we have are different from attitudes we would like to have, and sometimes these attitudes are the same. For your opinion of the following topics, please indicate the attitude you ACTUALLY have and the attitude you IDEALLY would like to have using the separate scales provided.” The items assessing actual attitudes were always presented before the items assessing desired attitudes. The participants always reported their actual and desired attitudes on single 7-point semantic differential scales ranging from 1 (extremely negative) to 7 (extremely positive). Participants reported the perceived stability of their actual and desired attitudes on a series of 7 point-scales anchored at
unstable-stable across different circumstances;
changing-permanent in time;
different-similar from now until 15 days later,
different-similar from now until 5 years later.
These four stability items were averaged to create an abstraction index for each type of attitude for each topic, with higher scores indicating higher perceived stability.
Keywords: Emotion regulation, Self-regulation, Motivation, Construal level
Background: In Study 1, we manipulated the instrumentality of anger, by using a hypothetical scenario that presents participants either with a goal that anger is likely to promote (i.e., confrontation) or with a goal that anger is unlikely to promote (i.e., collaboration). We predicted that participants in the high-level construal condition would be more motivated to experience anger when it is expected to be useful (i.e., when they need to confront) than when it is not. We did not expect the instrumentality of anger to influence participants in the low-level construal condition, since instrumental motives should be less salient to them.
Method: Participants were randomly assigned to one of the construal level conditions and completed the manipulation.
High-construal level condition: P. were asked to consider a goal (e.g., find and maintain a healthy relationship), explain why they wish to pursue it by identifying a higher order goal, then explain why they pursue that higher order goal by presenting an even higher order goal, etc.
Low-construal level condition: P. were presented with the same goal, but asked to explain how they wish to pursue it by identifying a lower order means, then explain how they pursue that lower order means, etc. Participants were then presented with a role-playing task. Participants were asked to play the role of a tenant whose refrigerator had been broken for over a week and their landlord has been avoiding them. Now, they are getting ready to speak with their landlord about fixing the fridge. Participants were then randomly assigned to one of two anger instrumentality conditions.
Participants in the high anger instrumentality condition were presented with a more confrontational goal: They were told that the lease is about to end and they have no plans to renew it.
Participants in the low anger instrumentality condition were presented with a more collaborative goal: They were told they needed to renew the lease with the landlord. Participants completed the emotional preferences measure, and rated both their current emotions and the goals they wanted to achieve during the interaction with their landlord.
Keywords: Explanation, comparison, categorisation, learning
Hypothesis: Experiment 1 evaluated the effects of prompts to explain and of prompts to compare on how people learn novel categories from examples, with a focus on the comparison of pairs of individual robots within the same category. We predicted that prompts to explain or to compare would be beneficial, that they would lead to more self-reported comparison, and that comparison would in turn be associated with positive learning outcomes.
Experiment 1: Method: The stimuli were pictures of eight robots, four robots (A-D) were labeled “Glorp rob”)ots” and four robots (E-H) were labeled “Drent robots.” There were four rules that could be used to categorize robots as either Glorp robots or Drent robots.
Foot rule
Antenna rule
Body-shape rule
Elbows/knees rule
At the beginning of the study phase, participants were told that they would study a set of robots and then answer questions about how to decide whether robots are Glorp robots or Drent robots:
Comparison prompts only Condition (“What are the similarities and differences between Glorp [Drent] robot X and Glorp [Drent] robot Y?”)
Explanation prompts only condition (“Try to explain why robot X is a Glorp [Drent] robot.”)
Both comparison and explanation prompts condition (“explain A, explain B, compare A and B”)
Control Condition (“Write out your thoughts below as you learn to categorize Glorp [Drent] robot X.”)
After studying the robots in one of these four ways, participants advanced to the rule-reporting phase. Participants were told, “We’re interested in any patterns that you noticed that might help differentiate Glorps and Drents.” Participants’ responses were evaluated by a coder choosing “yes” or “no” for each rule After completing the rule-reporting phase, participants answered self-report questions about the extent to which they engaged in explanation and in comparison (7-point scale).
“Did you notice yourself explaining what makes particular robots Glorp robots or Drent robots when the image of the eight robots was on-screen?”
“Did you notice yourself making comparisons between pairs of Glorp robots and pairs of Drent robots when the image of the eight robots was on-screen?”
Keywords: Bilingualism, Mental imagery, Foreign language
Hypothesis: We hypothesized that the use of a foreign language would diminish the vividness of mental imagery and that this would have consequences for choice.
Experiment 1: Method: Participants were randomly assigned to complete the experiment in the native language (here: English) or the foreign language (here: Spanish). Participants were then asked to try and mentally simulate 35 different sensory experiences across seven modalities that can be described as visual (e.g. “the sun as it is sinking below the horizon”), auditory (e.g. “the clapping of hands in applause”), tactile (e.g. “sand”), kinesthetic (e.g. “running upstairs”), gustatory (e.g. “salt”), olfactory (e.g. “fresh paint”), and organic (e.g. “a sore throat”). After each statement, participants were asked to rate the vividness of their mental simulations as one of the following options which were coded as 1 through 7: “no image”, “very vague and dim”, “vague and dim”, “not clear, but recognizable”, “moderately clear”, “very clear” or “perfectly clear”. Those in the foreign language condition additionally had an option to select “do not understand” for each item. Any item that received this response was excluded from that participant’s analysis.
Keywords: Gestalt, aesthetic perception, appreciation of art
Background: In this paper we show that viewers’ appreciation of cubist paintings is closely linked to the viewers’ ability to identify an object, or a Gestalt, from partial clues.
Method: Stimuli consisted of photographs of 120 cubist artworks by Pablo Picasso (47), Georges Braque (33), and Juan Gris (40). The study was structured in two blocks, each showing the stimuli in a randomized order. During the first block, subjects had to rate the pictures on liking. During the second block, participants rated how well they could detect objects within the artwork. All ratings were chosen from a 7-point-Likert-scale from 1 (‘not at all’) to 7 (‘very’). The results show that Gestalt formation is closely linked to appreciation; viewers much preferred paintings in which they were able to decipher concealed objects.
Keywords: Aesthetics, Individual differences, Artwork, Architecture, Landscapes, Faces
Background: In the first experiment, preferences were measured separately for individual classes of images: faces, natural scenes, interior and exterior architecture, and artwork. In Experiment 1, we computed preference agreement within separate groups of participants who each viewed only one stimulus domain. The participants performed two different tasks-a rating task designed to measure aesthetic appreciation and a “keypress” task designed to measure the amount of effort a person is willing to exert to view an image. We predicted that the more “natural” aesthetic domains (faces and natural landscapes) would show higher agreement than architecture or artwork, which are artifacts of human culture. We found highest agreement for faces, followed by natural landscapes, and then lower agreement for both interior and exterior architecture, and finally, lowest agreement for visual artworks.
Method: Rating task. Participants were asked to rate how aesthetically pleasing each image was, using a 7-point rating scale. Instructions were designed to encourage responses based on a participant’s own aesthetic evaluation. Participants were instructed to make judgments based on how much they, individually, were “moved” by the images, and that they should imagine that their evaluations of the images would be used to help curate a collection of images in the given domain. Key-press task. The viewing time task was used to gauge subjective reward value of the images. Observers were allowed to control the amount of time each image was displayed in a procedure similar to that of Aharon et al. (2001). Participants were told they could control how long they spent looking at each image by pressing keys to increase or decrease viewing time, and that they should look only as long as they enjoyed doing so.
Keywords: Facial likeness; familiarity; iconic images; face averages; prototype; mental representation
Background: In Experiment 1, we addressed the relationship between familiarity with a celebrity and likeness ratings given to images of them, as well as the idiosyncratic nature of likeness ratings. We examined individual and group variance in ratings (Hönekopp, 2006) to investigate the idiosyncratic nature of the perception of likeness. In addition, participants completed a speeded name verification task. If the perception of likeness from images maps on to the mental representation of identities, then it is reasonable to hypothesise that the higher the likeness rating, the faster that image will be verified as picturing the named identity.
Method: We used 15 images of each of 5 male Hollywood celebrities (Brad Pitt, Hugh Jackman, Matt Damon, Tom Cruise, and Tom Hanks). Participants began with a familiarity measure in which they were shown the name of each celebrity and asked to rate how familiar they were with each person on a scale from 1 (not very familiar) to 7 (very familiar). In the ratings task, participants viewed all 75 images in a random order, and were asked to indicate how good a likeness of that celebrity each image was. The likeness rating used a 1 to 7 scale from very bad likeness to very good likeness. After rating each image once, participants had a short break before rating each image a second time (again, in a novel random order). Participants also rated the images for trustworthiness and dominance, the results of which are not presented here. Following the ratings task, participants completed a speeded name verification task (30 trials). A name was displayed on the screen for 1500 ms, and then replaced by an image which remained on the screen until response. On half of the trials, the face matched the name. Participants responded via button press, indicating whether the image showed the same person as the name or not, and were instructed to respond as quickly and as accurately as possible.
Keywords: Trait inferences, Person perception, Configural processing
Background: Here, we therefore examined the role of configural processing in inferences of dominance and trustworthiness from people’s faces. We predicted that inverting faces would disrupt the perception of traits considered uniquely human (e.g., trustworthiness) but not the perception of traits believed to be shared by humans and animals (e.g., dominance).
Method: In Study 1, we randomly assigned participants to rate the trustworthiness or dominance of individual male faces. Participants rated each face both upright and inverted. This allowed us to calculate correlations for each participant’s ratings of upright and inverted presentations of the same identity, and thus test how much judgments of upright faces corresponded to judgments of inverted faces. Consistent with our hypothesis that configural processing plays a greater role in inferences of human-specific traits, we predicted that the correlation between upright and inverted faces would be stronger for the dominance than trustworthiness ratings.
Keywords: Cognitive Control; Task Switching; Emotion; Affective Priming; Negative Affect
Hypothesis/Background: Consistent with the recent theoretical frameworks mentioned above, and the idea that people avoid task contexts with high amounts of task switching. However, there is also reason to believe task switching could be experienced as positive. Experiment 1 served as a first, bidirectional test of the affective signature of task switching: task alternation primes could result in more positive or more negative judgments of neutral targets relative to task repetitions.
Method: In the task switching paradigm, target stimuli consisted of 320 high-frequent words each presented only once. All words could be judged on whether they were living or non-living, and whether they were smaller or larger than a basketball (see also, Braem, 2017; Schneider, 2015). Both dimensions were crossed orthogonally resulting in four lists of 80 words (see OSF). The Dutch words for SWITCH, CHANGE, REPEAT and AGAIN (WISSEL, VERANDER HERHAAL, OPNIEUW) were used as transition cues. Participants had to alternate between tasks when they saw the words SWITCH or CHANGE and had to repeat the task when they saw the words REPEAT or AGAIN. Each combination of transition cue and task was equiprobable. On the first trial of each block, participants were shown which task to start with (ALIVE? or SIZE?). If participants made a mistake, they would receive the message “WRONG!” along with the task of the erroneous trial (ALIVE? or SIZE?). In the affective priming procedure, target stimuli consisted of 60 Chinese pictographs. Participants had to evaluate these targets as rather negative or rather positive with the opposite hand that was used during task switching (D for negative, F for positive; or J for negative, K for positive). The transition cues functioned as prime stimuli. The Stimulus Onset Asynchrony (SOA) between the prime and the target was varied because it is a potential moderator of the priming effect (Fritz & Dreisbach, 2015), and could be 200, 400 or 800 ms. Each of the prime - SOA combinations were equiprobable, and each of the Chinese pictographs was paired once with each transition cue. To ensure processing of the prime stimuli, catch trials were dispersed throughout the task: When participant saw the word “PAPER” (PAPIER), they had to press space bar instead of evaluating the target.
Keywords: ATOM; compatibility; congruency; stimulus size; response position; SNARC
Background: In Experiment 1, size was the relevant stimulus feature and we varied the S-R mapping within participants. Results revealed a strong compatibility effect: Performance was better with the compatible mapping (small-left and large-right) than with the incompatible mapping (large-left and small-right). The results of our experiments suggest a strong relationship between the cognitive representation of physical (stimulus) size and response location in right-handers. The findings support the notion of a general magnitude code, as proposed in ATOM (theory of magnitude).
Method: In Experiment 1 we investigated the relationship between physical stimulus size and horizontal response position with a classic S-R compatibility task. Therefore, stimulus size (small vs. large) was the relevant stimulus feature and participants responded with two S-R mapping conditions in different blocks. In the compatible mapping condition, the small stimulus required a left-hand response, whereas the large stimulus required a right-hand response. In the incompatible mapping condition, the small stimulus required a right-hand response, and the large stimulus required a left-hand response. Our aim was to conceptually replicate a previous demonstration of the stimulus size - response location compatibility effect by Ren et al. (2011, Experiment 2), and to determine whether the effect would again be more pronounced for the right-hand response.
*Keywords: bandit problems; decision under uncertainty; belief learning; generalization; information foraging
Background: We introduce the spatially correlated multi-armed bandit as a paradigm for studying how people use generalization to guide search in larger problem spaces than traditionally used for studying human behaviour. We show that participants solve the exploration–exploitation dilemma by optimistically inflating expectations of reward by the underlying uncertainty, with recoverable evidence for the separate phenomena of directed (towards reducing uncer- tainty) and undirected (noisy) exploration.
Method: (Experiment 1) Participants searched for rewards on a 1 × 30 grid world, in which each tile represented a reward-generating arm of the bandit (Fig. 1a). The mean rewards of each tile were spatially correlated, with stronger correlations in smooth than in rough environments (between subjects; Fig. 1b). Participants were either assigned the goal of accumulating the largest average reward (accu- mulation condition), thereby balancing exploration–exploitation, or of finding the best overall tile (maximization condition), an explora- tion goal directed towards finding the global maximum. In addition, the search horizons (that is, number of clicks) alternated between rounds (within subject; short = 5 versus long = 10), with the order counterbalanced between subjects. We hypothesized that if function learning guides search behaviour, participants would perform better and learn faster in smooth environments, in which stronger spatial correlations reveal more information about nearby tiles.