Top-Rated Free Essay
Preview

Applying Classical Conditioning Toward the Physiological Detection of Concealed Information: Beyond Native Responses

Better Essays
9800 Words
Grammar
Grammar
Plagiarism
Plagiarism
Writing
Writing
Score
Score
Applying Classical Conditioning Toward the Physiological Detection of Concealed Information: Beyond Native Responses
Applying Classical Conditioning Toward the Physiological Detection of Concealed Information: Beyond Native Responses
Derek C. Tucker 6/8/2005

Psychology today is predominately concerned with phenomena which occur, “on average,” given a particular set of circumstances. Technology, however, is constantly forced to look deeper into phenomena that occur, “on average,” in order to improve the reliability of an instrument for whatever task the technology is to be used. With instruments such as the polygraph, that are applied to assess psychological states of an individual for a specific stimulus, the most technically reliable instrument is functional limited by the accuracy of the underlying psychological constructs for predicting physiological outcomes not “on average,” but “this time.”
While no psychological construct that I’m aware of reports 100% correlation with a specific physiological response, the next best thing is to be able to recognize when an event will likely be part of the average, so that one can modify the expectations for what might occur “this time”. This project investigates the potential of using classical conditioning to increase the reliability of “lie detectors” by using a paradigm that prevents people for whom the test will not work from undergoing critical (or relevant) question judgment (or classification). An issue with deception or lying, however, is that its definition is often subjective. It is imperative that the interrogator and the examinee both agree upon the interpretation of the stimuli used during the investigation. From the perspective of a fly, the Venus Fly Trap is behaving deceptively, though from the plant’s point of view, everything is as it should be. Thus, should a fly be equipped with a polygraph, if it were to ask, “are you a deceiver of flies,” or, “do you contain dangerous chemicals ” would be less effective than, “do you contain chemical X.”
In our society, rather than awaiting natural species adaptation, specialization, and differentiation, the “problem” of deception has been countered by the advance of lie detection techniques. Interestingly, many of these approaches implicitly rely on principles of classical conditioning for the generation and evaluation of results. According to the Committee to Review the Scientific Evidence for the Polygraph (2003), the underlying principle of the polygraph’s function under most paradigms relies on a lifetime of having episodes of intentional deception paired with fear and anxiety. In the parlance of classical conditioning, the unconditioned stimulus would presumably be the fear and anxiety that occur when one is afraid of being caught, while the conditioned stimulus would be the act of lying, or concealing information (Iacono, 2000). The unconditioned and conditioned responses, however, would vary by the individual experience . The original polygraph exam assumed a conditioned fear response as indicated by an increase of blood pressure that occurs on average (Martson, 1938), but as noted in Iacono (2000, p. 775): “There is no unique physiological response associated with lying, and there is no known physical substrate underlying what these tests measure. In fact, it is not clear what psychological processes are tapped by the techniques employed in polygraphy – or even how important deception per se is to their outcome. Polygraph operators are taught that their procedures most likely depend on a subject’s fear of the consequences of being detected. However, little research has been directed at this issue; it remains possible that other psychological constructs, such as the guilt and anxiety associated with lying or belief that a test works, are important. Because the psychological underpinnings of applied polygraphy are so poorly understood, it should be no surprise that the physical substrate underlying these techniques has received virtually no attention beyond the level of identifying useful peripheral measures.” Given the muddy waters of attempting to identify native conditioned responses to intentional deception, this project focuses on the potential to condition reflexive blink responses to statement veracity instead. That is, instead of “asking the plant” if it is a deceiver of flies or contains dangerous chemicals, this project asks whether the plant contains “chemical X,” a question whose answer should not be subjective The traditional polygraph would seek to either force a confession from the Venus Fly Trap during the interrogation, or determine deception based on increased sympathetic nervous system activity that occurred during a denial to a critical question which had been conditioned to occur after a lifetime of being scared when concealing information. The current approach, rather, installs and verifies a blink response contingent on statement veracity (regardless of the behavioral response) such that the outcome of the procedure would be the veracity of a given statement, not the instance of deception.

The Technology in Under Investigation
There are numerous technologies that are applied to the physiological detection of deception (PDD) which are often all combined under the category of “lie detectors.” This categorical title, however, is often a misnomer when applied to the actual function of the devices so named. This label is valid to the extent that if an individual being interrogated by a lie detector fails the exam, they have essentially divulged information which they would have otherwise concealed. This label does not mean that the technology functions by tapping into a cryptic signal specifically emitted during a “lie”. In fact, lie detection techniques vary in the specific measures being recorded, and by the structural design of the interrogation. Indeed, measures from brain waves, vocalizations, cardiovascular changes, and the thermal energy profile of the face have all been applied in efforts to reveal information a person would prefer to conceal.
The wide array of techniques that have been developed does not derive purely from a desire for diversity, but highlights the growing sentiment that new technology can overcome problems in the traditional polygraph. The recent report by the National Research Council on the scientific basis and validity of the polygraph, which also examined related measurement approaches (e.g., voice analysis, demeanor analysis, brain imaging), found that the accuracy of the polygraph test was better than chance but well below perfection (Committee to Review the Scientific Evidence on the Polygraph, 2003). The report adds that if a decision threshold is set to detect concealed information with any sensitivity, reliance on the polygraph test leads to high rates of false alarm, particularly when the incidence of deception is low. For this reason, the report concluded that the polygraph was an “unacceptable choice for DOE employee security screening” (Committee to Review the Scientific Evidence on the Polygraph, 2003 p. 219). Acknowledging this conclusion, the Department of Energy has pursued changes to its policy, however it maintains reliance on the polygraph for use in employee security screening citing that it cannot afford the risk of retiring the polygraph, although efforts to use the polygraph in conjunction with other information is a priority (McSlarrow, 2003). In sum, the technological assistance in detecting concealed information continues to be regarded as essential, and the status quo is not optimal.
Although the current techniques vary considerably, all are based on the measurement of “a naturally occurring physiological response or profile of responses that not only differentiates known deceptive from truthful answers but also allows accurate classification of answers as deceptive or truthful” (Committee to Review the Scientific Evidence on the Polygraph, 2003, p. 80). Importantly, the physiological responses are presumed to be elicited naturally in individuals who are concealing information. The next section illustrates the major approaches in which native responses have been used in the studies of the PDD. Following this review, an alternate approach in which classical conditioning is used to construct and verify a de novo association between a physiological response and knowledge that a statement is false is presented. “Polygraph testing is based on the presumptions that deception and truthfulness reliably elicit different psychological states across examinees and that physiological reactions differ reliably across examinees as a function of those psychological states” (Committee to Review the Scientific Evidence on the Polygraph, 2003, p. 71). The interrogation procedures of the polygraph presume either the psychological states of fear or anxiety that are typically induced during deception, or the orienting response which follows the recognition of a known stimulus (Committee to Review the Scientific Evidence on the Polygraph, 2003; Iacono, 2000). Changes in arousal or orientation lead to cardiovascular and respiratory responses such as pulse rate, blood pressure, respiration rate, perspiration, and peripheral vasomotor activity that correspond uniquely to the question about which an examinee wishes to conceal information (i.e., “relevant questions”).
The alterations in measured parameters during relevant questions, which are repeated several times to assure reliability, can be compared to alterations during non-relevant questions to estimate the likelihood that the examinee is being deceptive or intentionally concealing information (Marston, 1938; Larson, 1921; Lykken 1988). The first polygraph variation, the Relevant-Irrelevant Test (RIT), pitted questions relevant to the object of interrogation against questions that were completely irrelevant. For instance, during a murder investigation a relevant question may be, “Did you kill Jane Doe?” whereas an irrelevant question may be, “Were you born in Jerusalem?” Examiners compare the physiological responses to relevant questions to irrelevant questions. If relevant questions show greater responses, the examinee is classified as deceptive. A confound to this design is that relevant questions often contain an accusatory tone or specific stimuli, such as the word “kill,” that could elicit a response regardless of guilt, while irrelevant questions do not. As a remedy for this problem, a procedure known as the Control Question Test (CQT) was created (Reid, 1947).
Instead of comparing relevant questions directly to irrelevant questions, the CQT compares relevant questions to control questions designed to match the raw arousing aspects of the relevant questions. For instance, “Did you kill Jane Doe,” may be compared to, “Did you fondle little Jane.” Examiners look for greater responses to relevant questions than control to establish guilt, and equal or less response to relevant than control questions to establish innocence (Reid, 1947; Raskin, 1989).
Both the RIT and CQT formats rely on the fear, stress, or anxiety induced by the relevant questions to determine the examinee’s classification (Iacono, 2000). Other formats rely on the orienting response, such as the Peak of Tension Test, or the Guilty Knowledge Test (GKT), sometimes referred to as the Concealed Information Test. This design compares the responses to various completions of a relevant question such as, “Was the number of cars stolen . . . 2 . . . 4 . . . 6 . . . 8?” If the completion corresponding to knowledge that only the perpetrator would know elicits the greatest arousal, a deceptive classification is derived. The RIT, CQT, and GKT make up the most widely studied and applied PDD variations to date (Iacono, 2000). Although deriving better interrogation procedures can increase polygraph validity, the actual measures taken also contribute to critical reviews. One reason for the concern is that autonomic measures are multiply determined; they change as a consequence of various psychological and physical states; for instance, the orienting response is known to habituate (Ohman, Hamm, & Hugdahl, 2000). In consequence, the response from a deceptive examinee to a relevant item is expected to diminish during the repetitions embedded in the exam protocol (Committee to Review the Scientific Evidence on the Polygraph, 2003). Likewise, relevant questions may arouse an examinee in ways unanticipated during the design of the test that may remain uncorrected for at the termination of the exam (Iacono, 2000). The underlying difficulty is that existing approaches classify based on signals that are not always specific for deception or orientation (Committee to Review the Scientific Evidence on the Polygraph, 2003).
A different issue concerning polygraph measurement validity lies in the hardware from which the measures are taken. For instance, the strain gauge used to measure respiration is vulnerable to movement artifacts that can jeopardize the ability to compare responses to questions that are separated in time (Committee to Review the Scientific Evidence on the Polygraph, 2003). Most apparently, the very act of having sensors applied may unpredictably alter the psychological state of the examinee, and therefore alter signal validity. Many of the concerns raised by the polygraph’s hardware may be addressed by emerging remote measures such as thermography and voice stress analysis (VSA) which have been investigated with respect to the PDD.
Along with removing unwanted effects elicited by salient polygrahpy measures, remote measures may be used at mass screening sites such as transportation hubs, and for covert examinations. However, the limited studies on remote measures have suggested that upper limit of validity is no higher than the polygraph when used without simultaneous polygraph measurements (Department of Defense, 2002; Pavlidis, Eberhardt, and Levine, 2002). Although remote measures diminish confounds introduced by the polygraph examination itself, the responses measured remain subject to the multiple determinates that compromise signal specificity (Committee to Review the Scientific Evidence on the Polygraph, 2003).
In an effort to corner a more specific signal of deception, other approaches have incorporated the advance of brain function measurement technologies. ERPs have the longest history of investigation with respect to the PDD. Some ERP evidence suggests great improvement over the polygraph (Farwell & Smith, 2001), while others call into question the generalizability of reported results (Committee to Review the Scientific Evidence on the Polygraph, 2003), or demonstrate its greater than expected vulnerability to countermeasures (Rosenfeld et al, 2004). For now, the predominate opinion is that interrogation procedures incorporating ERPs perform with validity rivaling that of the traditional polygraph, but do not demonstrate a clear-cut advantage (Rosenfeld, 1991; United States General Accounting Office, 2001; Committee to Review the Scientific Evidence on the Polygraph, 2003).
Other brain measures like the fMRI have the capacity to suggest neural pathways underlying psychological states. Some researchers have investigated candidates of the pathway indicative of deception (Langleben, et al 2001; Spence, et al 2001; Lee et al, 2002). Acknowledging that deception presents through various psychological strategies, other researchers have demonstrated disparate pathways for different types of deception (Gannis et al, 2003). Still, other investigators have reported that despite significant results at the group level, the classification of individuals reveals an accuracy level similar to that of the polygraph (Kozel & Padgett, 2004). As it stands, the promise, costs, and complexity of fMRI with regard to the PDD contribute to its status as a continued focus of research, but restrain the widespread application to the field (Committee to Review the Scientific Evidence on the Polygraph, 2003). One theory behind the etiology of autonomic responses that correlate with deceptive intent suggests these responses have been conditioned to occur with deception. Following a lifetime of natural pairings of fear, stress, and anxiety with attempted lies or concealment of knowledge, deceptive intent elicits the response of fear, stress, or anxiety (Committee to Review the Scientific Evidence on the Polygraph, 2003). The major innovation of our approach is that we eliminate the assumption of conditioning over the lifetime. Instead of measuring a naturally emitted response whose extrapolation is assumed to reveal a psychological state, we train each examinee through a classical conditioning procedure to create a reflexive response indicative of a specific psychological state. That is, we induce a reflexive response contingent upon the veracity of a statement.
Pilot studies leading up to the present investigation indicate that having been classically conditioned does not eliminate the role of individual variation in the specific parameters of the conditioned response. Kozel (2004) suggested, with regard to the individual classification of examinees, that the large degree of idiosyncratic variation would be better accounted for if the thresholds that determined classifications were derived from an algorithm designed to establish the appropriate signal for an individual, instead of relying on an arbitrary highly significant number. Although that statement was made with regard to fMRI analysis, it can be applied to any analysis for an individual from measures that correlate to, yet are not uniquely specific for, the signal of importance. The present approach follows this suggestion. The target signal is first verified as differentially signifying false from true statements. Once an examinee is verified as showing a signal specific for false statements, the response parameters derived during the assessment of signal differentiability are used to subsequently prescribe the algorithm with which relevant questions will be judged.
It is possible that after an examinee has been through training, the assessment of the response may reveal that their conditioned response is not suitable to differentiate unknown items reliably. In fact, this outcome embodies the definition of the approach: by having the ability to assess signal specificity prior to critical question analysis, the validity of the classification based on thresholds and parameters derived from this assessment will be greater than the validity of a classification based on thresholds and parameters derived from group results. That is, this approach may lead to increased judgment validity by screening out individuals who do not elicit the differential response, thus limiting the impact of invalid assumptions. Once an examinee has been shown to elicit the differentiating response, the classification of a critical item may still be derived as ambiguous, though in a way that distinguishes the ambiguity from a generalized procedural confound. Thus, by using differential conditioning to generate and assess baseline, assisted by the removal of examinees impervious to the procedure, increased accuracy may be achieved in the independent classification of critical questions.

The Science in Question, Toward a Functional Integration with the Technology
Classical Conditioning is a diverse and deeply investigated field that began its’ scientific inquiry with Ian Pavlov in 1904 during a speculative Nobel Prize address (Lavond & Steinmetz, 2002), and has since expanded to underlie principles in neural network modeling, and other diverse topics likely unforeseen by Pavlov himself (Domjan, 2005). The most essential components that must be included in any discussion of classical conditioning are the Unconditioned Stimulus (US or UCS), which is a stimulus that upon presentation elicits a reflex response, the Unconditioned Response (UR); and Conditioned Stimulus (CS), which upon repeated pairings with the US comes to elicit a Conditioned Response (CR), which is typically, but not necessarily similar to the UR (see Table 1).
Component Abbreviation Operalization in this Study
Unconditioned Stimulus: A stimulus that naturally induces a reflexive response. US or UCS Air-puff to the eye
Unconditioned Response: The reflexive response induced by a natural stimulus. UR or UCR Blink
Conditioned Stimulus: A stimulus that is associated with the Unconditioned Stimulus CS The Veracity of a Statement
Conditioned Response: The learned response that occurs as a result of pairings between the Unconditioned Stimulus and the Conditioned Stimulus CR Blink
Inter-Stimulus-Interval: The time between the onset of the Conditioned Stimulus and the Unconditioned Stimulus ISI 495 ms
Inter-Trial-Interval: The time between conditioned stimulus presentations ITI Mean = 12s
Table 1. Summary of Essential Components of Classical Conditioning

Classical conditioning procedures involve a series of trials, often not all of which involve the coupling of the UCS with the CS. Typically, trails in which the CS is paired with the UCS are termed “CS+” trials, while those in which the CS is not paired with the UCS are called “CS-,” trials (Lavond & Steinmetz, 2002). In this study, the CS+ is a false statement, and the CS- is a true statement. The current document will depart from the standard convention with regards to the definition of a CS+, in that “CS+” will always refer to a false statement regardless of whether it is followed by the US (air puff) or not. Furthermore, all analysis presented besides the manipulation check, will only involve trials not paired with the air-puff. Although many classical conditioning experiments use CSs such as a flash of light, or a tone, some have ventured into using more complex semantic stimuli. Of greatest relevance is the finding that when differentially conditioned with an air-puff to the display of an incorrect math problem and not to correct math problems, participants took up the contingencies very well, and indeed reflexively blinked to incorrect math problems while not to correct ones (Grant, 1972). The current study is strikingly similar, requiring the substitution of math problems with right and wrong answers with objectively true and false statements. Thus, the currently presented paradigm is not a “lie detector,” but a “veracity detector,” and builds on the study presented by Grant (!972) by extending it from numerical semantic stimuli to verbal semantic stimuli. Interestingly, when pairing the air-puff with correct math problems and not to incorrect ones, the conditioning was much weaker (Grant, 1972). This is consistent with Dojaman’s (2005) point about ecological constraints on conditioned stimuli. He argues convincingly that conditioned stimuli with developmental relevance are far more effective than those with no apparent connection with the unconditioned stimulus. To that end, Grant (1972) argues that the reason conditioning wrong math problems but not right ones to the air-puff is due to the developmental history of being rewarded, not punished, for correct answers, and punished in some regard for incorrect ones. Following the suggestions of the literature, this study pairs the air-puff with false statements, rather than true, although an empirical investigation of any differences in the affinities of veracity for conditioning may be an interesting field for investigation in the future. The paradigm used in this study is applying classical conditioning for use in lie detection procedures. To that end, the contingencies trained on known true and false statements must generalize to unknown statements which would be the correlate of critical or relevant questions. Generalization of a conditioned response is the phenomenon whereby a conditioned response occurs to a stimulus that has not been specifically trained in the past, yet has some similarity to those stimuli which have been trained (Dayan & Abbot, 2001). However, as noted in Lerner & Orloff (1976), the operalizations of extinction and generalization are more alike than distinct, and may be confused in practice. This confusion arises because they both require a CS to be presented without a contingent US, after the CS-US contingency has been learned (Dayan & Abbot, 2001). Extinction studies examine to see how many trials are required before no CRs occur, while generalization studies examine to see if a CR occurs to a stimulus that is somehow similar, but not identical to the trained CS+ (Dayan & Abbot, 2001; Lerner & Orloff, 1976; Lavond & Steinmetz, 2002). It is important to distinguish between two types of generalization that may occur during this procedure. The unwanted generalization that can occur is from the CS+ to the CS-. With the example the current study, this would mean that reflexive blinks would occur to both true and false statements. If this generalization occurs, though, this will be detected by the analyses that assess the effectiveness of differential conditioning. The end result of a persistence of this generalization would be the inability to confidently judge the veracity of a critical stimulus for that individual. A second type of generalization is required for the application of classical conditioning to the “lie detection” field. Specifically, false statements that have not been reinforced or presented previously, should be able to elicit the same conditioned response that false statements which have been reinforced would elicit. Similarly, true statements that have not previously been presented should elicit the same response pattern as true statements that have been previously presented. The end result of a failure to obtain this second type of generalization would also be the inability to confidently judge the veracity of a critical stimulus . Importantly, the conditioned response pattern demonstrated in the differential conditioning of eye-blinks to correct vs. incorrect math problems was demonstrated with “C” responses, and not specifically for “V” responses. These two categories of response type are derived from their hypothetical etiologies of being truly conditioned (C) and voluntary (V) blink responses, and correspond to differences in the actual waveform of the recorded blink (Grant, 1972). The two types of response types are associated with different affinities for conditioned stimuli and different cognitive capacities. For instance, Grant (1972) argues that the V-type response pattern is associated with the ability or requirement to engage in complex, non-automatic associative learning, while the C-type response pattern represents closely associated contingencies and minimal executive influence. Since the successful differential conditioning to right and wrong arithmetic problems was found with C-type response patterns, it is feasible that the same response pattern may be found when using true and false statements. Alternatively, if determining statement veracity is more complex than mathematical correctness, the volitional V-type response pattern may be seen. This issue is relevant to the application of classical conditioning to the PDD in that volitional contamination of the reported conditioned responses would confuse the ability to reliably attribute a response to a conditioned reflex or an intentional action. V-type blinks often have a greater amplitude than related C-type blinks (Grant, 1972), a distinction that will be used in generating an algorithm that aims to counter any volitional contamination of V responses with C responses. It is possible, that depending on the specific statement, the V or C response may predominate given differential strengths in semantic expectancies generated from the beginning of a statement to its completion (Neely, 1990). To this end, each critical statement will be evaluated independently should evidence of differential conditioning to statement veracity arise. Investigating any idiosyncratic differences between statement conditionability is an important practical question for the usefulness of conditioning for lie detection purposes, and an important psychological question concerning neural processing of information (Grant, 1972). Most previous eye-blink conditioning studies use the Electromyogram (EMG) over the obicularis occuli muscle to assess blinks (Lavond & Steinmetz, 2002; Ghericke, Ornitz & Siddarth, 2002). The Vertical Electro Occulogram (VEOG), however, is used to detect blinks during such procedures as Electro Encepholography (EEG) in order to remove the impacts of blinks as an artifact of the recordings of brain function (Blumenthal et al, 2005). When comparing amplitudes and not latencies, EMG and VEOG have comparable sensitivity to the occurrence of a blink (Blumenthal et al, 2005; Ghericke, Ornitz & Siddarth, 2002). However, EMG measures the responsiveness of the orbicularis occuli, while VEOG is more sensitive to the levator palpebrae, two relatively independent muscles both required for instantiation of a blink (Blumenthal et al, 2005; Haines, 1999). Since both measures report different aspects of the same phenomenon, though the EMG is linked more closely to reflexive muscle response while the VEOG is linked more closely to the behavioral outcome (Blumenthal et al, 2005), it is possible that when recording both measures, one may be able to empirically, rather than by observationally, distinguish V-type responses from C-type responses. The hypothesis forwarded here is that C-type responses would show in the EMG measure independently of the VEOG measure, while V-type responses should not show outcome differences between the EMG and VEOG. Preliminary Original Idiosyncratic Algorithm Design
In order to judge each critical question for each individual participant, an algorithm was created in order to make the judgments systematic and objective. The algorithm is digital, and functions by determining the %CRs for a set of CSs using the same method whether the CS set is for the critical CS+ or CS-. The first step in the algorithm is to validate the reliability of a given measure (EMG and/or VEOG), that is to verify that the measure should be useful for the analysis of critical items by its ability to differentiate CS+s from CS-s using the Assessment items. Assessment items are trials which occur after the Training of differential conditioning and are drawn from the same stimulus pool as were trained. Besides allowing objective judgment of the veracity of an “unknown” critical statement via the blink responses to that item, another purpose of this algorithm is to screen out contamination of volitional responses during the critical questions which may occur due to a participant’s desire to please the experimenter, or from an attempt at countermeasures. A feature of the algorithm is that it prioritizes the EMG and VEOG measures, and only one measure is used to derive the ultimate score. When both measures are available, priority level 2 is considered first, and if not ambiguous, decides the fate of the critical item. Should priority level 2 be ambiguous, the measure with priority level 1 decides the fate of the critical item (see figure 3). The philosophy behind this difference concerns the idiosyncratic differences that individuals may possess with respect to their skeletal motor system and/or electrode placement (Tassinary, & Cacioppo, 2000; Blumenthal et al, 2005) such that for a given individual, one measure may be more appropriate for the digital analysis than another, and to the extent the measure’s responses have different origins, one may more reliably reflect true reflexive responses from volitional ones. Put differently, the digital algorithm assumes volitional or spontaneous action may be occurring, and under this circumstance VEOG and EMG responses are likely not to be as well correlated as they are when all responses are reflexive (Blumenthal et al, 2005), an assumption implicitly made in most eye-blink conditioning experiments.

Differential Conditioning Terminology
Differential conditioning uses as its’ dependent measure the difference between the percentage of CRs that occur during a CS+ trials and CS- trials, and is known as the “Percent Differential CRs” (Clark & Squire, 1999), and will be abbreviated in this paper as “%Diff.” This measure, as alluded to, is the difference between two other reportable measures, the “Percent Crs,” for the CS+ and CS-. These individual measures will be referred to in this paper as the “%CR+,” and “%CR-,” respectively. During the judgment of critical stimuli, the nature of the stimulus as a CS+ or CS- is theoretically unknown, thusly, the percent CRs to the critical stimuli will be presented as “CRX Determining the #CR is obviously a critical element in determining the %CR. A CR is typically determined to have occurred during a trial when there is a response amplitude increase, within a specified detection window, above a calculated baseline (Lavond & Steinmetz, 2002). Since the response amplitudes in this study have been standardized for each participant, the following characterizations of response amplitudes fit the profile of different responses: amplitudes 0 or less are “non-CRs,” amplitudes between 0 and 1 are “CRs,” and amplitudes above 1 are “Volitional.” By these definitions, %CRs can be calculated in two modes, either including volitional responses (V-type) with CRs, or deleting the entire trial with a volitional response, thus reducing the denominator, but preserving the measure as one of reflexive training (C-type). The %CR calculated that includes the “volitional” CRs as valid CRs is called the “Raw Score.” The %CR calculated that removes “volitional” CRs is called the “Corrected Score.” Likewise, the %Diffs for the two types of %CRs are called the “Raw %Diff,” and, “Corrected %Diff,” respectively. To this end, corrected %CRs are expected to reflect only the C-type responses, and the raw %CR is expected to reflect both the V-type and C-type responses. Because the digital algorithm has the responsibility of recognizing volitional contamination of the responses, whether from a motivation by the participant to please the examiner, or an attempt at countermeasures, the critical questions are examined both independently, and within their “companion set,” determined by the stem. That is, “Used for hugging” with the completions of “arm,” and, “foot,” are both analyzed with respect to operations that appear to occur as a function of the stem (“Used for Hugging”) independently from the operations of the statement veracity to screen for volitional responses. In the event no response pattern was generated as a function of the stem itself, the critical items of the companion set are considered separately.

Step 1: Verify Measures with Full Assessment Corrected %Diff

In order for a measure to be verified with the digital algorithm, it must past at least two steps. The first requirement is that the %Diff must be greater than 20%. This comes from the fact that in this study, each critical stimulus is presented 5 times, therefore, the minimum change between %differentials is 1/5 = 0.20 = 20%. Thus regardless if a significant difference is evident, the measure is only sensitive to changes of 20%. The second requirement is that the %CR+ must be above 25%, while the %CR- must be below 25%. The reason for this is requirement also is connected to the fact that critical items were presented 5 times. During the Test phase, the percentage reinforcement was 50%. In accordance with the Rescorla-Wagner model (Dayan & Abbot, 2001, see appendix F), the probability that a CR will occur to a CS is proportional to the percentage reinforcement. As demonstrated in Appendix F, the probability of a CR during a period of 50% reinforcement oscillates roughly between 40 & 60%. Using corrected measures, however, and the fact that critical items are novel in the course of the experiment while the items used to derive the Assessment phase have all been trained, it is feasible that an orienting response (Ohman, Hamm, & Hugdahl, 2000) occur to one of the five critical presentations, causing the response amplitude to be greater than 1.0 standard units, that trial would be deleted. If one trial were deleted, it can not be known whether that trial would have been a CR that was contaminated by the orienting response, or a non-response. Consequently, with the four remaining critical trials, if only one CR were to occur, the %CRX would be 25%. If the deleted trial response were indeed a CR, the actual %CRX would be 40%, within the expected range for a CS+ judgment. If the deleted trial was completely an artifact of the orienting response, the actual %CRX would be 20%, below the expected range for a CS+ judgment, and therefore qualifying for a CR- judgment. Therefore, concerning the Assessment phase %Diff, the known true (CS-) trials must show less than a 25% CR, and the known false (CS+) trials must show a %CR greater than 25%. This reflects the necessity to be able to reliably use derivations from the Assessment phase to extrapolate meaning from the responses to the critical items. Measures passing both requirements are considered, “Verified: A.” Despite the argument just presented, in this study, if the Corrected Assessment %Diff were greater than 50%, the measure was considered, “Verified: B,” for exploratory purposes.
Step 2: Sort Verified Measures If two measures are verified, that with the higher Corrected Full Assessment %Diff receives level 2 status, while the other receives level 1. If the %Diff between the two is 5% or less, the Corrected Assessment %Diff scores are used, with the measure obtaining the highest Corrected Assessment %Diff retaining level 2 status, and the one with the lower %Diff attaining level 1. In the study described, no tie breaks had to go beyond this point. The algorithm would continue, however, in the event of a tie in the Assessment Corrected scores, the Full-Assessment Raw scores would be used, followed by (in the event of another tie) the Assessment Raw scores.

Figure 1: Flow Chart Deriving the Verification of a Measure

Step 3: Judge Critical Stimuli

If the corrected CRX does not fall into an ambiguous case (see below), then the digital judgment is considered confidently screened for volitional responses, and receives a “2.” If the CRX > 25%, the final score is, “2+,” while if the CRX < 25%, the final score is, “2-.” If the corrected CRX falls into an ambiguous case, and the measure being used is of priority level 2, then the corrected CRX from the priority level 1 measure is used in the same way. If the corrected CRX of a priority level 1 measure is being used, the critical item may still be given a digital judgment, but will not be considered as having been successfully screened for volitional contamination, and the possible judgments will be between 1 and -1. If the raw CRX > 25%, the judgment is given of “1+,” while if the raw CRX < 25%, the judgment is given of, “1-.” If the raw scores fall into an ambiguous case (see below), the judgment given is “0.” For a description of ambiguous cases, see appendix E. Figure 6: Digital Algorithm Flow Chart for Judging CRX

Theoretical Predictions of the Current Study
1. Manipulation Check: The false statements (CS+UCS) that have been paired with an air-puff are expected to show greater blink responses than temporally related true statements (CS-pUCS) that were not paired with an air-puff.
2. Comparison of EMG and VEOG measure results: The %CRs generated by the EMG are expected to be the same as the %CRs generated by the VEOG. This will hypothesis will be tested by comparing the %CRs from both measures, and for both corrected and raw scores. Since the %CRs from the corrected scores are expected to reflect C-type responses alone, if there is a difference between the two measures for the corrected but not the raw scores, this would suggest that the corrected and raw scores do indeed reflect different types of responses, and will require investigation into the source of the difference. Importantly, if EMG and VEOG corrected measures are highly correlated, the response is expected to be of the C-form, whereas if they are not highly correlated, the response is expected to be of the V-form.
3. Replication of Differential Conditioning to right vs. wrong: False statements that were not paired with the air-puff (CS+), but had been trained previously, are expected to show greater conditioned blink responses than temporally related true statements (CS-), which also were not paired with the air-puff. Should this result not occur, we should rethink the assumed similarities between mathematic correctness and literary correctness. This hypothesis is tested by comparing the %CR+ to the %CR- for the Assessment items. If the %CR+ is significantly greater than the %CR-, the null hypothesis that the CR will occur equally to true and false statements after conditioning can be rejected.
4. Generalization of differential conditioned responding with veracity as a CS to novel statements: The novel statements with a false completion are expected to show greater conditioned blink responses than the novel statements with a true completion, none of which were paired with the air-puff. Should this not occur, aspects of the methodology should be questioned. This hypothesis is tested by comparing the %CR+ to the %CR- for the Critical-pool items. If the %CR+ is significantly greater than the %CR-, the null hypothesis that the CR will occur equally to novel true and false statements after conditioning can be rejected.
5. Investigation of differential statement conditionability and volitional contamination: The %CRs for both corrected and raw scores are expected to be similar for both critical items. Any difference suggests that the cognitive processes involved with specific statements may differ depending on their contents. If the corrected score is a better predictor for one over the other, it is expected that this critical item relies more on the C-type response, implying more automatic processing than the other. Similarly, if the raw score is a better predictor for one critical item over the other, it is expected that this item requires less and more executive processing to determine statement veracity. Evidence of differential processing may also be found if there are differences between the critical items in the comparison of EMG vs. VEOG. If EMG is better at differentiating the true and false statements of a particular stem while VEOG is better for the other stem, it is expected that the one preferring EMG is more automatically processed, and the one preferring VEOG requires more executive functioning.
6. Exploration of idiosyncratic results: The derived algorithm is expected to extrapolate the veracity of 4 novel statements according to their blink responses. This extrapolation will be based on response patterns generated by un-reinforced previously trained statements. Participants not generating a differential response pattern to previously trained true and false statements are not expected to show the differential response pattern to the novel statements. Should the algorithm improperly judge the statements of participants that generated differential responses, or properly judge the statements of participants that did not generate differential responses, the validity of the algorithm should be seriously questioned.

Methods
Participants and Design
Nineteen right-handed male participants, ages 18-25 (mean age 20.83 years), in good physical health and fluent in English were recruited from the University of Chicago. Participants’ task followed written informed consent, and their entire time in the lab was approximately 2 hours. Demographic information including recent alcohol, nicotine, herbal and prescription medication, history of illness and injury, and history of familial disorders was taken. Participants were compensated at the rate of five dollars per half hour for their participation in this study. Appendix A contains a summary of demographic information, appendix B describes exclusion criterion of subjects not included in the analysis (N=7).
Stimulus Materials
The UCS was a 5 psi air-puff that lasted 75 ms. and was delivered near the lateral corner of their left eye. The participants were fitted with safety goggles with a hole drilled through the protective glass through which tubing was attached. There were various levels of holes available to properly place the air puff for differently sized faces.
The Conditioned Stimuli (CS) were the veracities of a statement which followed the final word that completed an obviously true or false statement. Each statement was delivered through headphones as a digitally recorded sound file. A fully representative sample of statements was presented to each participant prior to the procedure to verify agreement on the assumptions of being true or false.

Since the property of statement veracity is not a component of the stimulus itself, but of the semantically derived abstraction from the configuration of the stimulus, using an elemental stimulus association paradigm, where objective stimuli are repeatedly paired with the US, is not suitable for this purpose. Lober and Lachnit (2002) demonstrated that configural learning is demonstrable using classical conditioning in humans by incorporating a biconditional discrimination design. This design extracts a particular configuration as a stimulus rather than the raw stimuli themselves. Stimuli are always presented as pairs, which in this study will be represented by a stem (the start of a statement) and a completion (the end of a statement), the schematic for biconditional discrimination is: AB+, CD+, AD-, CD- (Lober & Lachnit, 2002). The first part of any pair is either A or C. For this study, A and C correspond to such statement stems as, “When heated melts,” and, “When heated burns,” respectively. The second part of the pair is either B or D. For this study, they correspond to such completions as, “wood,” and “ice,” respectively. Thusly, the combinations AB+ and CD+ would correspond to, “When heated melts: wood,” and, “When heated burns: ice,” respectively, both are false statements, and would be the CS+. Similarly, the configurations AD- and CD- would correspond to “When heated burns: wood,” and, “When heated melts: ice,” both true statements, corresponding to the CS-. Therefore, the difference between a CS+ and CS- has nothing to do with correlations between the raw stimuli that are presented and the US, but instead the configuration of stimuli that are presented, which, for this study, cause statements to be either true or false. Not surprisingly, there is evidence for differential neural pathways with respect to the hippocampus, cerebellum, and amygdala between elemental associative learning, and biconditional discrimination (Good, de Hoz, & Morris 1998; Choi & Moore, 2003; Fanselow & Poulos, 2005), which would likely come into play should this paradigm be extended to neural imaging measures.
To further underscore this contingency, a screen appeared after the completion showing “TRUE” and “FALSE” in a green and red box, respectively, with the side of appearance (right or left) for each box varying randomly. The left most box corresponded to the “F” key, and the rightmost box corresponded to the “J” key. Participants were to respond by pressing “F” or “J” corresponding to the left or right appearance of the correct response, “True” or “False” which varied randomly. The response screen randomization restrictions protected against more than 3 consecutive “F” or “J” responses, thus avoiding a possible confound by preserving the requirement of attention to predict accuracy (see Table 2).
Phase Number of Trials Design of Stems Design of Trials
Training 120 (60 true stimuli, 60 false stimuli) 12 stems (1-12) 90% False reinforced
Assessment 60 (30 true stimuli, 30 false stimuli) 6 stems (3, 4, 7, 8, 9 & 10) 50% False reinforced
Test 120 (60 true stimuli, 60 false stimuli) 12 stems (old stems 2-11; critical stems 13 & 14 ) 50% False reinforced Experimental control, visual, and audio presentation were performed by a custom program developed in the E-Prime environment from Psychology Software Tools, Inc., on a PC running the Windows 98SE Operating System. Eye movement and blink signal were measured using EMG over the right (contralateral to the air-puff) orbicularis occuli (OOC), and VEOG on the left (ipsilateral to the air-puff) eye. The tubing in the safety goggles fed back to a custom built computer controlled system to calibrate and time the air flow. Output of the EMG, VEOG, and air puff mechanism was recorded on a second
Table 2 – Phase Design

Windows 98SE PC through the Acknowledge © program, and their veracity judgment was stored as a text file through E-Prime.
Design
We incorporated the differential eye-blink conditioning paradigm into a three phase procedure: Training, Assessment, and Test. The conditioned stimulus was the veracity of the statement, which presumably was recognized subsequent to the completion’s presentation. The unconditioned stimulus (UCS: air-puff) was delivered 745 ms. following the onset of the completion, which never overlapped or preceded the completion’s articulation (see figure 3).
Figure 3: Generic Trial Timing

One of the trickiest parts of employing a classical conditioning paradigm with a semantic configural CS is determining the time of onset for the CS. Theoretically, the CS is the realization of statement veracity, which occurs sometime after the appearance of the statement completion. The precision of CS onset is important in determining the ISI, which is of critical importance for the interpretation of classical conditioning data (Herbert, Eckerman & Stanton, 2003). In humans, ISIs around 500ms appear to be most effective for conditioning (Lavond & Steinmetz, 2002). In order to achieve this, the US must be presented around 500ms after the expected time of realization of the statement veracity. Although neural processing contingent on the presence of a visual object in the environment occurs in ERP studies as early as 100ms (Josh Correll, personal communication; Montoya et al, 1996), and can be influenced by conditioning (Montoya et al, 1996), the semantic associations related to the visual object induce their earliest effect between 200 (Josh Correll, personal communication; Arecchi, 2004; Bonte & Blomert, 2004) and 250ms (Martin-Loeches, Sommer, & Hinojosa, 2005). Therefore, it is reasonable to expect that the participant becomes aware of the veracity of the statement 250ms after the presentation of the completion, and a UCS should follow by 500 ms. In sum, 750ms should elapse between the onset of the completion and the administration of a US, thus corresponding to a 500ms ISI, when one takes into account the time for semantic processing. A “Response Screen” appeared 300ms following the termination of the time slot for the air-puff, indicating which button to press to indicate “true” or “false”. The participant’s response to the statement was included as a potential method to assess participant attention throughout the experiment, but the response itself is theoretically unimportant for the conditioning procedure. Preliminary report of response errors are presented in Appendices C & D. The “Response Screen” stayed on until the participant responded. An inter-trial interval (ITI) of eight to fourteen seconds with a mean of 12 seconds, was randomized throughout the experiment. The length of the mean ITI is exceedingly important because it determines the overall length of the study (Carrillo et al, 1997). Based on a study by Prokasy, Grant, and Myers (1958), which showed that contingency acquisition increases with the mean ITI length, with most significant gains occurring up to 30 seconds, most human eye blink studies use a mean ITI between 20 and 30 seconds. Indeed, some believe that humans do not condition with ITI’s of less than 10 seconds, and many animal studies use ITI’s of several minutes (Lavond & Steinmetz, 2002). Carrillo et al (1997), however, argue that the practical gains from having 5, 10, or 15 second ITI’s may outweigh the theoretical gains from the relative increased contingency acquisition. One of the problems with longer ITI’s with human participants is that they often get bored or sleepy (Lavond & Steinmetz, 2002; Personal Observation). This is of particular concern when a Trace paradigm is being used, as the level of awareness for stimulus contingencies has a direct effect on contingency acquisition (Clark & Squire, 1999). Background tasks are often given to human participants to maintain attentiveness during the ITI (Lavond & Steinmetz, 2002), but depending on the task, may interfere with acquisition (Papka, Ivry & Woodruff-Pak, 1995) in a manner that interacts with personality variables between participants (Tracy et al, 1999). In this study, we used music clips to fill all but the final second of the ITI , and have records of participant behavioral response errors, which offer some indication of attentiveness.

True stimuli that were temporally correlated over the course of the experiment with false stimuli that received an air-puff were termed “CS-pUCS” for analysis purposes. False stimuli that were reinforced were identified as, “CS+UCS” False stimuli that were not reinforced were identified as “CS+” while true stimuli temporally correlated with them were identified as “CS-.”
Phase Design
The field of Classical Conditioning has many other components and phenomena that must be considered in designing a paradigm. One critical variable is the percentage of reinforcement (the frequency that CSs are contingent with the US). In addition, phenomena such as blocking and generalization/extinction (discussed below) have to be considered when constructing the design of a classical conditioning experiment. Of theoretical importance, the introduction of classical conditioning paradigms to the “lie detection” field will have to, at some point, take into account countermeasures incorporated by individuals who have investigated the technology, and specifically trained against it. Thusly, the psychological implications of classical conditioning, and the ecological constraints of using it as a “lie detector” are important considerations.. Estimating ecological concerns, the current paradigm has been divided into 3 phases: Training, Assessment, and Test. During the Training phase, 120 stimuli were presented, 60 true and 60 false. According to Levond and Steinmetz (2002), humans usually take between 25 and 50 trials in order to learn the association between a CS and US. To be conservative, we assumed it would take 60 trials to create a reliable response. Therefore, since each CS+ trial was paired with a CS- trial, 120 trials were continuously presented in the Training Phase. Of the false stimuli, 90% (54) were reinforced with an air-puff. The Training phase is included to “teach” the participant the differential contingency between false statements and true statements with the air-puff. Thus, the Training phase should have a high percentage reinforcement to promote CR specificity and prevent generalization of the CR from the false (CS+) to true (CS-) statements. Considering the biconditional discrimination format and the potential for participants to loose focus with long ITI’s, the generalization from false to true statements is a serious threat. As such, high percentage reinforcement for the Training phase is essential.

During the Assessment phase, 60 stimuli were presented, 30 true and 30 false. Of the false stimuli, 50% (15) were reinforced. The Testing phase was designed to appear as two Assessment phases, 120 stimuli, 60 true, 60 false, 30 false stimuli were reinforced. The additional element in the Test phase was that critical items that had never before been seen during the procedure, and were never reinforced, were embedded, thus allowing an analysis of the detection of the veracity of the statement. The “true” critical stimuli were the stem + completions: “Used for hugging – arm,” and “Used for kicking – foot.” The “false” critical stimuli were “Used for hugging – foot,” and “Used for kicking – arm”. The stem, “Used for hugging,” is referred to in the analysis as stem “M,” or, “13,” while, “Used for kicking,” is referred to as stem “N,” or, “14.” The Assessment phase is included to assess the contingencies learned in the Training phase, and to assure that CRs will occur to un-reinforced CS+, while not to an analogous CS-s. As demonstrated in Appendix F, assuring generalizability requires lower percentage reinforcement than does assuring acquisition. Since the probability of a CR gravitates to the percentage of reinforcement (Dayan & Abbot, 2001; Lerner & Orloff, 1976), and the %CR to CS- (true) statements is expected to be near 0.00, the Assessment phase percentage reinforcement was set at 50%, a value which should enable distinguishable distributions between %CS+ and %CS- responses, as well as permit generalization to novel stimuli. Theoretically, one would not proceed to the Test phase unless the Assessment phase items produced CR patterns that could distinguish between true and false statements. The Test phase, as previously stated, should be identical to the Assessment phase, except for the inclusion of the “critical” or “relevant” items (statements) which are being judged or evaluated. The critical items receive no reinforcement, but provided lack of extinction and a high probability of generalization, they should elicit similar percent %CRs as the items in the Assessment phase.

The Dependent measures were the presence of a blink and its’ magnitude, as detected by vertical electro-oculogram (VEOG) on the eye that was being presented with the UCS, and electromyograph (EMG) on the eye contralateral to the air-puff.
Procedure
Participants were greeted upon arrival at the lab by the experimenter who brought them into the room where preparation for EMG and VEOG recording was administered. After informed consent was obtained and demographics completed, participants were read a description of their task while the electrodes were being placed. The electrodes were affixed as described in Tassinary & Cacioppo (2000). The impedance between the electrodes was verified as less than 5 kiloohms. Next, participants were seated in an overstuffed chair in the testing chamber, reclined to 45°, fitted with the air-hose safety goggles, head phones, and given a cup of water. They were also given a remote keyboard on which they were to make the judgments to the statements. The experimenter then summarized the experimental procedure, reminding participants that air puffs would only be received if the completion following the stem lead to a false statement. That is, the conditioning contingency was made explicit to participants.
The experimenter was located in a separate control room monitoring the participant during program execution. Next, an automated experimental control presentation program was started that was used to verify a good signal from the EMG and VEOG, and asses normal blinks from the participant. After verification of setup, an adaptation program using 4 true stem-completion pairs was used to accustom the participant to the environment, and to make sure they had full comprehension of the task.
Any questions or concerns by the participant were addressed, and the Training – Assessment – Test sequence was initiated. The participants were allowed a break between each phase, and also between 60 trial segments of the Test phase during which time, the recordings were suspended.
Data Analysis
For each single trial, the peak detection window was set as 250 ms before the offset of CS to 200 ms after the offset of the UCS, in addition to the 75ms for the potential air-puff. Most classical conditioning studies consider a CR any activity above baseline that occurs within a window between the onset of the CS and the onset of the UCS; however, when analyzing trials on which no UCS was administered, it is common to double the length of the window to beyond the occurrence of the would be UCS in order to capture CRs with longer latencies (Lavond & Steinmetz, 2002). Since the detection window must avoid inclusion of the alpha response, or a reflexive blink to the CS, it is typically defined near the center of the ISI (Lavond & Steinmetz, 2002). As a consequence, our detection window was set to 250ms prior to the air-puff onset, and 200ms after the air-puff completion (a duration of 75 ms), for a total of 575 ms. The baseline was determined by applying a third order Butterworth digital high pass filter to the data for the cumulative the time interval starting 40 ms before the onset of CS until 300 ms after the onset of the CS. The peak magnitude was determined by subtracting the baseline value from the maximum value of the response that occurred within the peak detection window. Response magnitudes were standardized within participants to account for systematic individual differences in response patterns, in order to isolate the changes in response due to the experimental procedures (the analysis was also performed on non-standardized data with an equivalent pattern of results, and is available upon request from the authors). For the Training Phase, the first two reinforced false (CS+UCS) and yoked true (CS-pUCS) trials were not included in the analysis to account for differences in response due to the novelty of the US. Repeated measures analysis of variance (ANOVA) were performed to compare peak magnitudes of CS+ vs. CS-, CS+UCS vs. CS-pUCS, and critical CS+ vs. critical CS-. This analysis was performed on each phase independently. The Huynh-Feldt correction for rejection of sphericity was used, and an alpha of 0.05 was the threshold for significance. Comparisons concerning the CS+UCS and CS-pUCS were analyzed as a manipulation check, and are presented with the aforementioned analyses only.
The %CR+ and %CR- were calculated, using the same strategy as described in the digital algorithm for “Raw Scores,” and compared with a two tailed paired t-test. For these analyses, the Assessment phase CS+ and CS- populations were combined with the Test phase (non-critical) CS+ and CS- trials into a category called, “Full Assessment.” For exploratory purposes, the CS+ critical items and CS- critical items from both critical stems were pooled into a category called, “Critical Pool.” The %CR+ and %CR- were also calculated for the individual critical questions. The same analysis was performed for the corrected %CRs as described in the digital algorithm for “Corrected Scores.” A paired t-test was also performed for each analysis group between the VEOG and EMG to assess any systematic measure differences.
Results & Preliminary Discussion
1. Manipulation Check (CS+UCS vs. CS-pUCS) During the Training phase, there was a main effect of the air-puff on VEOG response (F[1,11] = 252.17 p

You May Also Find These Documents Helpful

  • Good Essays

    Here is a simple classical conditioning experiment that you can perform on yourself at home. You will need a bell (or something you can ring), a hand-held mirror, and a room that becomes completely dark when the light is turned off. Hold the bell while standing in the room near the light switch. Once in position, you should ring the bell and then immediately turn off the light. After waiting in total darkness for about 15 seconds, turn the light back on. Wait another 15 seconds with the light on, and then ring the bell and immediately turn the light back off (again waiting 15 seconds in the dark). Repeat this procedure 20 to 30 times, making sure that in each case the bell is rung immediately before the light is turned off. After numerous pairings, you should be ready to see the results. With the light on, watch your eyes closely in the mirror and then ring the bell. Your pupils should dilate slightly even without a change in light!…

    • 910 Words
    • 4 Pages
    Good Essays
  • Good Essays

    Learning refers to the process whereby experience produces a fairly lasting and adaptive change in behaviour (Passer et al., 2009). Classical conditioning is the process of learning by association which signals the approaching arrival of a significant event. It involves pairing a neutral stimulus with an unconditioned stimulus (US) that will elicit an unconditioned response (UR). With repeated pairings, the neutral stimulus becomes a conditioned stimulus (CS) that evokes a conditioned response (CR) similar to the original UR (Passer et al., 2009).…

    • 685 Words
    • 3 Pages
    Good Essays
  • Good Essays

    Classical conditioning is a type of learning in which a potent stimulus obtains the ability to evoke an innate response that was originally elicited by a neutral stimulus. In classical conditioning, a UR is an event that occurs naturally in response to some stimuli. On the other hand, a UR is the stimulus that naturally and automatically triggers a response without learning. A CS in classical conditioning is an originally neutral stimulus that, through learning, comes to be associated with some unlearned responses. Finally, a CR is the learned response to the originally neutral but now conditioned stimulus (CITE BOOK). These are the basic components involved in classical conditioning. Classical conditioning theory was first discovered and described…

    • 1000 Words
    • 4 Pages
    Good Essays
  • Satisfactory Essays

    Little Albert

    • 401 Words
    • 2 Pages

    Watson J. B., & Rayner, R. (1920). Conditioned emotional reactions. Journal of Experimental Psychology, 3(1), 1–14.…

    • 401 Words
    • 2 Pages
    Satisfactory Essays
  • Satisfactory Essays

    Classical Conditioning

    • 375 Words
    • 1 Page

    Classical conditioning is a method of conditioning in which associations are made between a natural stimulus and a learned, neutral stimulus. I consider classical conditioning to be very important because it’s such an efficient way of teaching, training or conditioning people or animals, especially children. Classical conditioning could be used for psychological distress like phobias. For example, Mary cover jones put a child with a fear of rabbits in a room with the rabbit far way. Then she gave him his favorite food and put the rabbit closer. Associating the pleasure of food with the feared object made him no longer scared of rabbits. This applies to my life because my mom used this method when she raised me. I was scared of riding my bike because I fell off it once. So every time I attempted pedaling she would give me a dollar. Finally she put 5 dollars all way down the street and told me to bike there and get it; making me lose my phobia of bikes.…

    • 375 Words
    • 1 Page
    Satisfactory Essays
  • Satisfactory Essays

    The behavioral perspective that exemplifies how external environmental events condition our observable behavior. People and animals behave as they normal would everyday due to their environment and past experiences. Scientific methodology takes up a huge part in behaviorism and how studies can be objectively measured. Our environment and what we are faced with daily affects our observable behavior (aka our response). When it comes to the behavioral perspective, there are two types of conditioning: (1) classical conditioning and (2) operant conditioning. Classical conditioning is a process in learning by association and determining what our behavior is. Operant conditioning is the process of learning by consequence and rewards. I believe operant conditioning is seen more commonly today with the parenting style for kids and the obedience process of animals. If a child is punished for wrong-doing, they likely will not do that task again. But if they are never told right from wrong, they will continue doing things that are frowned-upon. Operant conditioning is a very good way for parents and pet owners to teach right from wrong. Classical conditioning is good for…

    • 363 Words
    • 2 Pages
    Satisfactory Essays
  • Powerful Essays

    Detecting Deception

    • 5352 Words
    • 22 Pages

    Currently, the most successful and widespread system is the polygraph which monitors uncontrolled changes in heart rate and electro-dermal response, as a result of the subject’s arousal to deceit. Unfortunately, its widespread use does not necessarily mean it is a perfect system. Firstly, in order for it to take the necessary measurements, it needs to be continuously connected to the subject’s body. This means that the subject must be cooperative and in close proximity to the device. Secondly, it requires accurate calibration at the beginning of every session, so that a baseline of measurements can be established. Occasionally, it may still fail to give accurate readings, despite the calibration step; if for example, the subject’s heart rate increases for reasons unrelated to deception. Furthermore, the polygraph is an overt system, which means that the subject knows they are being monitored and also knows what measurements are being made.…

    • 5352 Words
    • 22 Pages
    Powerful Essays
  • Good Essays

    Raynor & Watson carried out a controversial experiment in 1920 using classical conditioning to try and understand the origins of different fears and phobias. They observed the behaviour of a boy named Albert and found that he took a liking to a white rat and did not demonstrate any fear when subjected to the rat; the only thing that he expressed any fear of was a loud noise which would make him cry. They combined the loud noise with the rat which he later developed a phobia of. Both experiments demonstrate the effects of classical conditioning.…

    • 802 Words
    • 4 Pages
    Good Essays
  • Satisfactory Essays

    Classical Conditioning

    • 868 Words
    • 4 Pages

    |1. Jamie was talked into riding on the roller coaster |Terror ride |Fear |Coaster |Fear/cold sweat |…

    • 868 Words
    • 4 Pages
    Satisfactory Essays
  • Good Essays

    Well, let’s start with ancient China. “In ancient China suspect were given rice to chew, the idea being that liar’s would be too nervous to salivate so the rice would remain dry”(Gaidos 38). Interesting, I wondered how that work out for them. But how do we know now? “When someone’s lying, subtle changes in the vocal cord are said to occur as a result of the stress, producing a distorted sound wave”(Prusher 1). When people get stressed they sound different I guess. I’m not sure how that works, but it’s not my problem. But there is a guy who is working on making an improved lie detector. Truster’s new lie detector concentrates mostly on stress levels in people's voice by measuring its “cognitive messages” and low-frequency waves that are not audible and are 85% accurate (Prusher 1). So stress is a bad thing whenever you’ve been accused of something. Good to know, I guess if you're truthful and never stressed this test will be a breeze. “Using a complex algorithm and nine different parameters, the inventors say they are able to pinpoint whether the person's stress is caused by lying, excitement, exaggeration, or an emotional conflict”(Prusher 1). Lot’s of big words but basically Truster’s is an inventor who is making a lie detector that is determined by stress…

    • 811 Words
    • 4 Pages
    Good Essays
  • Good Essays

    Class: Psychology of Personality Subjects the class covers are as follows: ***What is Personality? -the eight perspectives of personality -objective and subjective approach to personality assessment *…

    • 540 Words
    • 3 Pages
    Good Essays
  • Good Essays

    1920's Inventions

    • 740 Words
    • 3 Pages

    food. At that time a bandage consisted of separate gauze and adhesive tape that you would cut to size…

    • 740 Words
    • 3 Pages
    Good Essays
  • Good Essays

    Our understanding of classical conditioning, operant conditioning, and observational learning has allowed us to unlock many of the answers we sought to learn about human behavior. Classical conditioning is a technique of behavioral training, coined by Ivan Pavlov, which basically states that an organism learns through establishing associations between different events and stimuli. This helps us understand human behavior in an assortment of ways. It makes it clear that almost everything we do is based on patterns of stimulus and response. For example, if you were bitten aggressively by a dog as a child, you may be still scared of dogs today. That is because the dog caused you pain, which in turn caused you have anxiety towards dogs. Because you associated the dog with pain, and the pain caused you to have anxiety, therefore you brain associated seeing a dog with feelings of anxiety. Same thing applies to getting a text message. Let’s say you’re sitting around doing nothing an all of the sudden your phone vibrates. You’ll probably go and check to see what message you got. This relates to a classical conditioning experiment because you have associated your phone vibrating with getting a message.…

    • 658 Words
    • 3 Pages
    Good Essays
  • Powerful Essays

    Miss

    • 6631 Words
    • 27 Pages

    Rethinking the Benefits of ‘Justified’ Deception What are we to make of that unique practice associated with some psychology experiments, the intentional deception of the research subjects? Psychologists argue that they are not using malicious or garden-variety deception, but deception of the ‘justified’ kind. They are quick to assure critics that these subjects will endure minimal risks, if any, while participating. Indeed, some give the impression that there is too much fuss over deception: many of the ethical sermons being preached to social scientists seem to assume that those participating in research projects would never encounter given discomforts if they did not participate in the research . . . deceptive information is presented at every turn,…

    • 6631 Words
    • 27 Pages
    Powerful Essays