CS as an effect: action-based evaluative conditioning depends on temporal contiguity

ABSTRACT

In this study, we analysed Evaluative Conditioning (EC) with a recently introduced S-R paradigm [Blask et al., 2016 Blask, K. , Frings, C. , & Walther, E. (2016). Doing is for feeling. Journal of Experimental Psychology: General , 145 (10), 1263–1268. https://doi.org/ 10.1037/xge0000211 [Crossref], [PubMed], [Web of Science ®] , [Google Scholar]. Doing is for feeling. Journal of Experimental Psychology: General, 145(10), 1263–1268], in which valence first transfers from an affective US to a response and then from this valence-laden response to a novel CS. Based on action control research, the CS in this procedure could be interpreted as an effect of the response. Hence, temporal contiguity of the valence-laden response and the CS should determine whether EC occurs. We tested this hypothesis in an experiment, in which we varied the time interval between the response and the subsequent CS. Results show that EC effects occurred only under conditions of high temporal contiguity thus supporting our assumptions. This finding sheds light on the processes underlying action-based EC in particular and evaluative learning in general.     

Reevaluating evaluative conditioning: a nonassociative explanation of conditioning effects in the visual evaluative conditioning paradigm.

In 2 studies, the authors investigated whether evaluative conditioning (EC) is an associative phenomenon. Experiment 1 compared a standard EC paradigm with nonpaired and no-treatment control conditions. EC effects were obtained only when the conditioned stimulus (CS) and unconditioned stimulus (UCS) were rated as perceptually similar. However, similar EC effects were obtained in both control groups. An earlier failure to obtain EC effects was reanalyzed in Experiment 2. Conditioning-like effects were found when comparing a CS with the most perceptually similar UCSs used in the procedure but not when analyzing a CS rating with respect to the UCS with which it was paired during conditioning. The implications are that EC effects found in many studies are not due to associative learning and that the special characteristics of EC (conditioning without awareness and resistance to extinction) are probably nonassociative artifacts of the EC paradigm.     

Resistance to extinction in evaluative conditioning

A well-demonstrated phenomenon in traditional Pavlovian conditioning research with humans is that of experimental extinction. In contrast, human evaluative conditioning research suggests that evaluative learning shows marked resistance to extinction. Here, the authors replicate both findings concurrently. Two differential fear conditioning experiments with an electrocutaneous stimulus as the unconditioned stimulus evidenced (a) sensitivity to extinction using an autonomic skin-conductance measure and (b) complete resistance to extinction using an affective-priming measure. The results corroborate the idea that evaluative conditioning is more resistant to extinction than is expectancy learning (F. Baeyens, P. Eelen, & G. Crombez, 1995).     

Both trace and delay conditioning of evaluative responses depend on contingency awareness

Whereas previous evaluative conditioning (EC) studies produced inconsistent results concerning the role of contingency knowledge, there are classical eye-blink conditioning studies suggesting that declarative processes are involved in trace conditioning but not in delay conditioning. In two EC experiments pairing neutral sounds (conditioned stimuli; CSs) with affective pictures (unconditioned stimuli; USs), we tested whether the type of conditioning procedure affects the relationship between awareness and EC. Auditory CSs and visual USs were either overlapping and co-terminating (Delay Conditioning Group), or they were separated by a short temporal gap (Trace Conditioning Group). In both groups, contingency awareness was manipulated by varying the duration at which the US was presented during conditioning. Furthermore, in addition to a post-conditioning measure of awareness, US detectability was measured concurrently during the learning phase in Experiment 2. US exposure duration affected both measures of awareness. The data of both experiments demonstrate that EC does not occur if the USs are presented at very brief durations, but reliable EC effects can be observed at longer US exposure durations. As there were no differences between trace and delay conditioning, the data suggest that contingency awareness plays a crucial role in trace and in delay conditioning of evaluative responses. These results challenge automatic association-formation accounts of EC, but they are in line with ‘single-process’ accounts assuming EC to be dependent on conscious declarative knowledge about CS–US contingencies.     

Goal relevance moderates evaluative conditioning effects

An important process by which preferences emerge is evaluative conditioning, defined as a change in the evaluation of a stimulus by pairing it repeatedly and consistently with an affective stimulus. The current research focuses on the role of motivation in this learning process. Specifically, it was investigated whether a conditioning procedure that is relevant to an individual's current goals is more effective than an irrelevant procedure. To this end, beverages were conditioned with either disgusted faces (relevant) or fearful faces (irrelevant). The results showed that thirsty(rather than non-thirsty) participants’ choice and evaluation of beverages were influenced by pairing beverages with disgust but not with fear. As similar results were obtained under optimal and suboptimal presentation of the conditioned stimuli, it is suggested that goals can affect automatic, associative learning, adding to the emerging body of research demonstrating that goals unconsciously affect evaluative processes.     

Context effects in evaluative conditioning of implicit evaluations

Propositional models of evaluative conditioning postulate that the impact of stimulus pairings on liking should depend not on the pairings themselves but on what the pairings imply about the relation between stimuli. Hence, context manipulations that change the implications of stimulus pairings should moderate evaluative conditioning. We manipulated context by varying the way in which context cues were paired with affective outcomes while keeping the pairings between target cues and affective outcomes constant. All participants saw one target cue compound that was followed by a positive outcome (XF+) and another target cue compound that was followed by a negative outcome (YG?). In condition Same, each context cue was consistently paired with a positive or negative outcome, regardless of whether it was presented alone or in compound with another cue (A+, B+, AB+; C?, D?, CD?). In condition Opposite, however, a context cue was paired with a certain outcome when presented alone and with an outcome of the opposite valence when presented in a compound with another cue (A+, B+, AB?; C?, D?, CD+). Employing several implicit measures, we assessed the implicit evaluations of the target cues X and Y. In all three studies, the outcome of the measurement procedure differed between conditions. In condition Same, the positively paired cue X was evaluated more positively than the negatively paired cue Y. In condition Opposite, however, this preference was not present. This pattern of results suggests that EC is determined not only by the objective pairings but also by the context in which these pairings occur. Implications for models of evaluative conditioning are discussed.     

The role of CS-US contingency in human evaluative conditioning

Recent studies suggest that in Pavlovian conditioning, two different processes may be operative: signal learning and evaluative learning, resulting in two qualitatively different associative structures. Signal-learning is hypothesized to be responsible for providing us with genuine predictors (CS) for significant events (US). This proposition logically entails that the statistical-correlational relation, i.e. the contingency between CS and US should be a crucial determinant of signal learning. Evaluative conditioning, on the other hand, refers to the observation that the mere pairing of neutral with (dis)liked stimuli changes the valence of the originally neutral stimuli in a (negative) positive direction. As argued elsewhere, evaluative conditioning is probably based on the CS acquiring a mere referential value to the US, without any genuine CS-US expectancy being involved. From this, it was hypothesized that evaluative conditioning might not be dependent on CS-US contingency. Using the standard evaluative conditioning paradigm, four different levels of CS-US contingency were created on a between-subject base. The overall effect of evaluative conditioning was strongly significant, and was not mediated by awareness of the CS-US relation. Of crucial importance, this conditioning effect did not interact with the level of contingency, supporting the hypothesis that CS-US contingency is not a crucial determinant of evaluative conditioning. Moreover, this effect was obtained in a situation in which Ss simultaneously evidenced to have consciously registered quite accurately the different levels of CS-US contingency.     

Human evaluative conditioning: Order of stimulus presentation

In an experiment designed to demonstrate evaluative conditioning, subjects were shown 48 pictures of sculptures that they rated on a scale with 21 categories (?10 to +10). Then, the two most liked pictures (L) were paired with pictures from the categories ?1, 0, or +1 (N). In contrast to prior experiments, subjects were given either forward conditioning (N-L) or backward conditioning (L-N) trials but not both. Four other neutral stimuli were paired with each other (N-N) and acted as control stimulus pairs. After conditioning, the stimuli were rated a second time. There was a statistically significant difference in evaluative ratings showing a change of the evaluative tone of the previously neutral stimuli in a positive direction only after forward conditioning. This finding is inconsistent with results of prior experiments and challenges the assumption of Martin and Levey (1987) that evaluative conditioning is different from human classical conditioning.     

Measuring evaluative conditioning using the Implicit Association Test

When a neutral stimulus is paired with a stimulus that has strong affective properties, these properties often appear to be transferred to the neutral stimulus. This learning has been termed evaluative conditioning. In two experiments, participants first learned the ‘meanings’ of four non-words. Two of these meanings were affectively positive, and two were affectively negative. Transfer of affect was measured in the Implicit Association Test (IAT). In the IAT, a participant’s affective response to an item is inferred from his or her ability to categorise that item with other pleasant and unpleasant stimuli. An item easily categorised with other liked stimuli is deemed to be liked itself. In both experiments, evidence for transfer of affect was observed in the IAT. Thus, non-words given pleasant meanings in training (evaluative conditioning) were more easily categorised with pleasant than unpleasant personality characteristics, compared to non-words given unpleasant meanings. The results suggest that the IAT is a useful way to test for evaluative conditioning. It is both sensitive to the transfer of affective properties from one stimulus to another, and, because affect is measured indirectly, the results are unlikely to be due to the demand characteristics of the experiment.     

Observational evaluative conditioning of an embedded stimulus element

Evaluative conditioning refers to the observation that the mere paired presentation of a neutral stimulus (CS) with a liked or disliked stimulus (US) may result in the neutral stimulus itself acquiring positive or negative valence. In most studies, the CS is an autonomous, invariant stimulus, and the subject directly experiences both CS and US. In this experiment, we investigated whether evaluative conditioning can be extended to a situation wherein the CS is no more than an invariant element of a complex, variable stimulus configuration, and wherein the subject experiences the CS–US co-occurrences indirectly, i.e. by observing a socius who is exposed to the CS–US pairings and facially expresses either liking or disliking the US. During acquisition, subjects watched video-taped sequences of an actor drinking a glass containing a liquid and facially expressing either liking or disliking the drink. The stimulus element which was systematically paired with the actor's facial expression of liking or disliking, was whether the glass contained a ‘foot’ or no ‘foot’ (CS), while other characteristics of the scenes were systematically varied and paired equally often with an expression of like and dislike. Next, valence ratings were obtained for pictures in which the CS element (foot/no foot) was embedded. A clear observational evaluative learning effect could be demonstrated when the feature CS was embedded in objects identical to those presented during learning, but not when it was embedded in new objects. These data demonstrate the possibility of vicarious evaluative conditioning of an embedded stimulus element, but probably at a lower level of abstraction than intended.     

A conceptual and theoretical analysis of evaluative conditioning

Evaluative conditioning is best defined as an effect, that is, as a change in the valence of a stimulus that results from pairing the stimulus with another stimulus. This definition has several advantages that are made explicit in this paper. One of the advantages is that it clarifies that evaluative conditioning can be due to multiple processes. Therefore, the conditions under which evaluative conditioning is observed can depend on the processes that underlie a particular manifestation of evaluative conditioning. This could explain why there are so many conflicting results about the conditions under which evaluative conditioning can be found. Future research should adopt a meta-conditional approach that focuses not only on whether a certain condition is crucial for obtaining evaluative conditioning but should also examine when a certain condition is crucial.     

The effect of comparative context on evaluative conditioning

Evaluative conditioning (EC) is a change in the evaluation of a stimulus after the stimulus co-occurred with affective stimuli. The present research examined whether EC of one stimulus depends also on the co-occurrence of another stimulus with positive or negative stimuli. We paired two target people with affective stimuli. We found that a person who appeared eight times with positive stimuli and eight times with negative stimuli was liked more when the other person appeared always (16 times) with negative stimuli than when the other person appeared always with positive stimuli. The manipulation did not change the US evaluation or the general standard on the value dimension of what a positive or negative stimulus is. We suggest that like other evaluative traits (e.g., evil, pretty) co-occurrence with affective stimuli is sensitive to temporary standards. The manipulation changed the standard of what co-occurrence with affective stimuli is considered positive versus negative.     

Exploring evaluative conditioning using a working memory task

Johnsrude, Owen, Zhao, and White (1999) demonstrated that evaluative conditioning (EC) can occur even while participants conduct a working memory task. This is consistent with claims that EC can occur outside awareness, and hence is an automatic process. In their study, three abstract patterns were paired differentially with a candy reinforcer: 90, 50, or 10%. On reinforced trials a red ball was presented in one of three locations on a computer screen. Participants were instructed to keep a tally for each location. We argue that if EC is automatic, then it will not consume cognitive/attentional resources, and hence co-occurrence of red ball and reinforcer should be irrelevant. Results from four experiments indicate that learning is only evident when the red ball and reinforcer co-occur or when the complexity of the working memory task is reduced. Together, the results suggest that EC is governed by specific attentional constraints on working memory. This interpretation questions the automatic nature of EC.     

Successful differential evaluative conditioning using simultaneous and trace conditioning procedures in the picture-picture paradigm

This paper describes a study designed to investigate the efficacy of two traditional classical conditioning procedures in generating evaluative conditioning (EC) in the picture-picture paradigm in human participants. Differential EC was found using both simultaneous and trace conditioning procedures. In addition, the use of a block-subblock (BSB) nonpaired control condition and full counterbalancing of conditioned stimulus-unconditioned stimulus (CS-UCS) pairings across participants indicated that the observed EC effects were the result of associative learning. Examination of whether successful conditioning could take place with or without conscious awareness of CS-UCS contingencies was inconclusive. The results provide evidence for EC as an associative process and also provide some insight into the possible conditioning parameters that might successfully generate EC.     

Differential fear conditioning induces reciprocal changes in the sensory responses of lateral amygdala neurons to the CS(+) and CS(-)

In classical fear conditioning, a neutral sensory stimulus (CS) acquires the ability to elicit fear responses after pairing to a noxious unconditioned stimulus (US). As amygdala lesions prevent the acquisition of fear responses and the lateral amygdaloid (LA) nucleus is the main input station of the amygdala for auditory afferents, the effect of auditory fear conditioning on the sensory responsiveness of LA neurons has been examined. Although conditioning was shown to increase CS-evoked LA responses, the specificity of the changes in responsiveness was not tested. Because conditioning might induce nonspecific increases in LA responses to auditory afferents, we re-examined this issue in conscious, head-restrained cats using a differential conditioning paradigm where only one of two tones (CS(+) but not CS(-)) was paired to the US. Differential conditioning increased unit and field responses to the CS(+), whereas responses to the CS(-) decreased. Such changes have never been observed in the amygdala except in cases where the CS(-) had been paired to the US before and fear responses not extinguished. This suggests that fear conditioning is not only accompanied by potentiation of amygdalopetal pathways conveying the CS(+) but also by the depression of sensory inputs unpaired to noxious stimuli.     

Reducing automatically activated racial prejudice through implicit evaluative conditioning

The authors report a set of experiments that use an implicit evaluative conditioning procedure to reduce automatically activated racial prejudice in White participants in a short period and with relatively few trials. Experiment 1 demonstrated that participants were unaware of the repeated conditioned stimulus-unconditioned stimulus (CS-US) pairings of Black-good and White-bad. In Experiment 2, the procedure was found to be effective in reducing prejudice as indicated by an evaluative priming measure of automatically activated racial attitudes. In Experiment 3, this reduction in prejudice was found to persist throughout a 2-day separation between the conditioning procedure and the administration of the dependent measure. The implications of the present findings for the persistence of automatically activated racial prejudice are discussed.     

Human evaluative conditioning: acquisition trials, presentation schedule, evaluative style and contingency awareness.

Two different processes may be operative in human Pavlovian conditioning: signal learning and evaluative learning. Whereas most studies on evaluative conditioning focused on a mere demonstration of the phenomenon or on a theoretical analysis of the underlying processes, some basic parameters of evaluative learning are still unexplored. Hence, using the standard neutral picture--(dis)liked picture pairing paradigm (Baeyens, Eelen & Van den Bergh, 1990), in this study the effect of two parameters of evaluative conditioning was assessed on a between-subjects base, namely the Number of Acquisition Trials (2/5/10/20) and the Presentation Schedule of the stimulus pairs (blockwise or random). Additionally, the study included an exploratory analysis of the potential effects of the Evaluative Style of subjects (Feelers vs Thinkers, operationalized in terms of speed of emitting evaluations). Finally, the relationship between contingency awareness and evaluative learning was reassessed. Neutral-liked conditioning was found to be quadratically related to the number of acquisition trials (increase in effect up to 10 trials, decrease from 10 to 20 trials), whereas neutral-disliked conditioning linearly increased with increasing numbers of trials. Randomized vs blockwise presentation schedules of the stimulus pairs did differentially affect the overall pattern of conditioning, but in a way which was both unexpected and difficult to account for theoretically. Both the Evaluative Style of subjects and contingency awareness were demonstrated to be generally orthogonal to conditioned shifts in CS valence. Based on these findings, some practical suggestions are provided for the application of evaluating conditioning based therapeutical interventions to affective-behavioral disorders which are centred around inappropriate (dis)likes.