Using the Scharff-technique to elicit information: How to effectively establish the “illusion of knowing it all”?

Abstract

The Scharff-technique is used for eliciting information from human sources. At the very core of the technique is the “illusion of knowing it all” tactic, which aims to inflate a source's perception of how much knowledge an interviewer holds about the event to be discussed. For the current study, we mapped the effects following two different ways of introducing this particular tactic; a traditional way of implementation where the interviewer explicitly states that s/he already knows most of the important information (the traditional condition), and a new way of implementation where the interviewer just starts to present the information that s/he holds (the just start condition). The two versions were compared in two separate experiments. In Experiment 1 (N = 60), we measured the participants’ perceptions of the interviewer's knowledge, and in Experiment 2 (N = 60), the participants’ perceptions of the interviewer's knowledge gaps. We found that participants in the just start condition (a) believed the interviewer had more knowledge (Experiment 1), and (b) searched less actively for gaps in the interviewer's knowledge (Experiment 2), compared to the traditional condition. We will discuss the current findings and how sources test and perceive the knowledge his or her interviewer possesses within a framework of social hypothesis testing.

Resumen

La técnica de Scharff se utiliza para extraer información de fuentes humanas. En el meollo de la técnica está la táctica de la “ilusión de saberlo todo”, que apunta a engordar la percepción de una fuente sobre cuánto conocimiento posee un entrevistador sobre el hecho que se aborda. Para realizar este estudio cartografiamos los efectos derivados de la introducción de esta táctica particular, un método tradicional de aplicación, en el que el entrevistador afirma de modo explícito que ya conoce casi toda la información importante (la condición tradicional) y una manera nueva de implementación, en la que el entrevistador empieza a presentar la información que posee (la condición de simplemente iniciar la condición). Se comparó ambas versiones en dos experimentos distintos. En el experimento 1 (N = 60) medimos la percepción que tenían los participantes de los conocimientos del entrevistador y en el experimento 2 (N = 60) la percepción que tenían los participantes de las lagunas de conocimiento del entrevistador. Se halló que los participantes de la condición de “simplemente iniciar” (a) creían que el entrevistador poseía más conocimientos (experimento 1) y (b) buscaban de un modo menos activo las lagunas de conocimiento del entrevistador (experimento 2), en comparación con la condición “tradicional”. Comentaremos estos resultados y cómo perciben y ponen a prueba las fuentes los conocimientos de su entrevistador, en el marco de la prueba de hipótesis social.

Human intelligence (HUMINT) is “a category of intelligence derived from information collected and provided by human sources” (NATO, 2014, p. 115). Typically, in a HUMINT interaction an interviewer aims to collect information about past or future criminal activities (Evans, Meissner, Brandon, Russano, & Kleinman, 2010; Vrij & Granhag, 2014). A specific form of HUMINT gathering is information elicitation, for which the goal is to collect information in such a manner that the sources remain unaware of the true purpose of the interaction (Justice, Bhatt, Brandon, & Kleinman, 2010). The main aims are here that the sources underestimate how much new information they have revealed and remain unaware of the interviewer's information objectives.

Humans who strive toward goals develop strategies and plans (Fiske & Taylor, 1991, 2008). In the HUMINT context, sources often aim for a specific goal (e.g., money or protection from prosecution) and what they offer in return is information. Additionally, sources are often cooperative to some extent; they are willing to share some but not all information they hold. In order to pursue their goal, sources often use so-called counter-interrogation strategies; for example, “I will not say very much during the interrogation,” “I will try to figure out what they are after, and then make sure I do not give them what they want,” and “It is meaningless to deny or withhold what they already know” (Scharff, 1950; Soufan, 2011; Toliver, 1997). Recently, Alison et al. (2014) presented a field study in which they showed the relevance of such counter-interrogation strategies.

The Scharff-technique

The Scharff-technique aims to collect information from sources that are motivated to reveal some but not all information (Granhag, 2010). An interviewer taking the perspective of the source lies at the very core of the Scharff-technique. Perspective taking refers to the “cognitive capacity to consider the world from other viewpoints and allows an individual to anticipate the behavior and reactions of others” (Galinsky, Maddux, Gilin, & White, 2008, p. 378). Taking the perspective of others is effective in negotiations (Galinsky & Mussweiler, 2001), and of importance for criminal and HUMINT interviewers (Granhag & Hartwig, 2015; Soufan, 2011).

The Scharff-technique is a collection of tactics that draws on the interviewer's insights about the source's goals and counter-interrogation strategies (Granhag, 2010). The friendly approach tactic stipulates that the interviewer establishes and maintains a pleasant, conversational atmosphere during the interview. When employing the illusion of knowing it all tactic, the interviewer presents already known information, makes clear that s/he is well-informed regarding the topic to be discussed, and gives the source the opportunity to add details. The confirmation/disconfirmation tactic aims to elicit specific pieces of information as the interviewer presents claims that s/he seeks to have affirmed or negated. The not pressing for information tactic requires the interviewer to collect information by asking very few, if any, questions. Finally, using the ignore new information tactic means the interviewer conceals his or her interest for information and treats the information that the source reveals as known or unimportant (for more detailed descriptions on the Scharff tactics, see Granhag, Montecinos, & Oleszkiewicz, 2015; May, Granhag, & Oleszkiewicz, 2014).

In a series of studies, the Scharff-technique has been compared to the Direct Approach, which is a combination of open-ended and specific questions (US Army, 2006). In accordance with the Field Manual 2–22.3 (US Army, 2006) and the Executive Order No. 13941 (US Government, 2009), the Direct Approach is the most commonly used intelligence interviewing technique in the field (Redlich, Kelly, & Miller, 2011; Semel, 2013). All previous studies have used an experimental paradigm mirroring important features of a typical HUMINT interaction (Granhag et al., 2015a). Simply put, participants received incomplete information on a planned attack and were instructed to strike a balance between not revealing too much or too little information in a subsequent interview. In past studies, the Scharff-technique has outperformed the Direct Approach by all important measures. First, the Scharff-technique resulted in relatively more new information (e.g., May & Granhag, 2015; see Granhag, Oleszkiewicz, Strömwall, & Kleinman, 2015 for sources who varied in their levels of cooperation and knowledge). Second, the sources interviewed with the Scharff-technique underestimated how much new information they revealed, whereas the sources interviewed with the Direct Approach overestimated how much new information they revealed (e.g., May et al., 2014; Oleszkiewicz, Granhag, & Cancino Montecinos, 2014). Finally, the sources interviewed by the Scharff-technique found it relatively more difficult to read the interviewer's information objective (e.g., May et al., 2014; Oleszkiewicz, Granhag, & Kleinman, 2014).

In previous studies, the illusion of knowing it all tactic played an important role in terms of collecting new information. The presentation of known information (i.e., the illusion of knowing it all tactic) followed by an open-ended question resulted in more new information compared to simply asking an open-ended question (e.g., May et al., 2014; Oleszkiewicz et al., 2014a). That is, by presenting known information, the interviewer made the source to reveal information beyond what was disclosed by the interviewer. Another aim of the illusion of knowing it all tactic is to inflate the source's perception of how much knowledge the interviewer holds about the event. Specifically, the interviewer steers the source's focus towards his or her knowledge of the event and steers it away from his or her knowledge gaps. The current study is about the efficacy of different ways of introducing the illusion of knowing it all tactic aiming to inflate the source's perception of the interviewer's knowledge.

Sources Exploring the Interviewer's Knowledge

Humans are naturally goal-oriented (Aarts, 2012), and in order to decide if and how to pursue a goal, they form and test hypotheses. Trope and Liberman (1996) presented a framework for social hypothesis testing that consists of five steps and can be applied to a source that tests the amount and relevance of an interviewer's knowledge. At first, the source may formulate a hypothesis (e.g., “The interviewer holds some important information”). S/he may then derive if-then rules from stored knowledge in order to test the hypothesis (e.g., “If the interviewer holds important information, then s/he knows who founded the group”). In the third step the source searches for relevant information in his or her memory from past interviews or actively during an ongoing interaction with the interviewer in order to test these if-then rules (e.g., “The interviewer knows that a woman founded the group”). In accordance with these if-then rules, the source may then interpret and categorize the collected information (e.g., “A woman founded the group” is categorized as important knowledge), and finally draw an inference by assessing the likelihood that the hypothesis is true or not (e.g., “The interviewer actually holds some important knowledge”).

Trope and Liberman (1996) distinguished between two broad hypothesis-testing methods. The diagnostic testing refers to a comprehensive analysis of the hypothesis and its alternatives. In contrast, the pseudodiagnostic testing refers to the neglect of the alternative hypothesis; persons search for hypothesis-consistent information, interpret information and/or draw inferences in favor of the hypothesis. Since the latter processes often lead to a confirmation of the focal hypothesis, it is also considered a hypothesis confirmation strategy (Skov & Sherman, 1986), or is referred to as a positive test strategy, as it can, but does not have to, confirm the focal hypothesis (Hodgins & Zuckerman, 1993; Klayman & Ha, 1987). Studies show that humans generally prefer the diagnostic strategy to test social hypotheses (Bassok & Trope, 1984; Trope & Bassok, 1983), and this independent of whether or not they are instructed to test a specific hypothesis (Trope, Bassok, & Alon 1984). However, persons prefer the positive test strategy when presented with equally diagnostic hypothesis-consistent and alternative-consistent features (Devine, Hirt, & Gehrke, 1990; Skov & Sherman, 1986), especially when instructed to test an extreme (radical) hypothesis (e.g., “The interviewer holds all important information”) in comparison to a more moderate hypothesis (e.g., “The interviewer holds some important information”; Trope & Bassok, 1983). In brief, many interview situations (including HUMINT interactions) are characterized by the interviewer aiming at influencing a) the hypothesis that the source formulates about the interviewer's knowledge, and how the source tests this hypothesis, and b) what inference the source draws from this testing (Granhag & Hartwig, 2015; Soufan, 2011).

The Present Research

The current paper examines two ways of introducing the illusion of knowing it all tactic. For the traditional condition the interviewer began presenting information in a fashion similar to that found in previous studies (e.g., May & Granhag, 2015), stating explicitly: “I already posses most of the most important information and let me just share that information first.” In contrast, for the just start condition the interviewer simply presented the known information without an explicit statement regarding the amount or relevance of known information.

The present research examines these two conditions in two separate experiments. In Experiment 1 we measured how sources perceived and tested the interviewer's knowledge in terms of known information, and in Experiment 2 in terms of unknown information. We examined the different dependent variables in two studies in order to avoid order effects and confounds. In both studies the participants called an interviewer who then employed the illusion of knowing it all tactic and eventually interrupted the conversation. Participants were then asked to record their perceptions of the interview by filling out a questionnaire which differed between the two experiments.

Drawing on research on social hypothesis testing, we predicted that compared to sources in the traditional condition, sources in the just start condition would test the interviewer's knowledge (a) more actively for known information (Experiment 1), and (b) less for unknown information (Experiment 2). The reason for this was that the sources in the traditional condition were presented with the interviewer's statement to “possess most of the most important information.” In short, this statement may trigger alertness and we therefore expected the sources to test a comparatively more extreme (radical) hypothesis, thereby using a positive test strategy to uncover information gaps in the interviewer's knowledge. In contrast, the sources in the just start condition were not presented with a statement on the amount and relevance of the known information. Hence, we expected them to be surprised by all of the known information, and, accordingly, to test a comparatively more moderate hypothesis and focus more on the known information possessed by the interviewer. Consequently, we predicted that the sources in the just start condition would infer that the interviewer (a) held more knowledge (Experiment 1), and (b) had fewer knowledge gaps (Experiment 2), compared to the sources in the traditional condition. The rationale behind this was that in the traditional condition the interviewer gave the sources a point of reference for the amount of known information (“I already possess most of the most important information, so let me just share that information first”), and that they would focus more on the gaps in the interviewer's knowledge. In contrast, we expected that the sources in the just start condition would be unsuspecting regarding the amount of known information, and that they would search more actively for information that the interviewer possessed.

Experiment 1

For this experiment we measured the participants’ perceptions of the interviewer's knowledge. We predicted that the participants in the just start condition would write down comparatively more information that was known to the interviewer (Hypothesis 1), and would think comparatively more about further information that the interviewer could have (Hypothesis 2). Furthermore, we hypothesized that the participants in the just start condition would perceive that the interviewer held comparatively more information (Hypothesis 3).

Method

Participants

The study included 60 participants (49 female and 11 male; 42 psychology students, 16 other students, and two employees). The mean age was 23.55 years (SD = 6.29, ages ranging from 18 to 59), and they were randomly assigned to one of the two interview conditions (30 participants for each condition). The psychology students participated in exchange for partial course credit and the other participants received 10 Euros as compensation for their participation. Originally, 62 persons were recruited for this study, but two had to be excluded (one misunderstood the instructions, and one reported technical problems).

Materials

We produced a list containing of 14 themes on a terrorist group in order to prepare instructions for participants and protocols for the interviewers. Each theme consisted of two pieces of information with varying levels of specificity. For example, for the theme “date of the attack,” the general piece of information was “around Christmas,” and the specific piece of information was “the 27^th of December.”

Procedure

Background and planning. Participants received basic instructions and information concerning the group and their activities in order to prepare the role of a source calling a police contact. Specifically, they had to memorize the specific information on all 14 themes, as they were not allowed to have notes available during the phone call. The participants were instructed to strike a balance between not revealing too much information (because they sympathized with the group), and not revealing too little information (because they needed help from the police). Furthermore, they were allowed to provide false information to the interviewer, and were not informed if the police contact already possessed any information. After 20 minutes of preparation time the participants were asked to fill out a memory test about the 14 pieces of background information (e.g., “On what exact date will the bomb attack take place?”). An identical memory test was filled out immediately after the interview. These two tests were scored and participants were only included in the analysis if they achieved a score of 11 or higher on both the pre and post-interview tests. In Experiment 1, all participants fulfilled the criteria. After checking the correctness of this test, the experimenter left the room and the participants called the interviewer via phone.

The interview. Three interviewers conducted the interviews (two female and one male), and all of them held approximately the same amount of interviews in each condition. They were trained in conducting the interviews and instructed to closely follow the two interview protocols.

• (a)

  Just start condition. For the just start condition the interviewer made the phone call and started the interview by asking the source if everything was okay with the phone connection. Then the interviewer explained kindly that s/he might have to interrupt the conversation as s/he was expecting an important call from his or her boss. The interviewer continued by inquiring about the source's well-being. After the source finished answering, the interviewer showed understanding for the source's situation and explained that s/he knew about different options that could be used to help the source (friendly approach tactic). Then the interviewer employed the illusion of knowing it all tactic by stating the following: “I think it's really good that you are just contacting me now, so that I had a chance to familiarize myself with your situation, and also had time to think about MDA and their current planning” (MDA was the name of the terrorist group). The interviewer then presented information on seven themes with a general level of specificity (the name of the group, the number of members in general, their origins, the number of members planning the current attack, the extent of their knowledge on building bombs, how well the bomb could be concealed, and the location of the bomb's production). The interviewer ended his or her presentation by stating, “But I’m sorry, I think I should stop talking so much, so let's get back to the conditions of our deal and how you can contribute.” Next, the interviewer's phone rang loud enough for the source to hear and the interviewer explained that s/he had to take the phone call and would call the source back in 5 to 10 minutes. In reality, the interviewer did not call back.

• (b)

  Traditional condition. The interview protocols in the just start condition and the traditional condition differed solely with respect to the introduction of the illusion of knowing it all tactic. For the traditional condition the interviewer stated: “I think it's really good that you are just contacting me now, so that I had time to familiarize myself with your situation. After thinking through the whole thing on MDA again, I’m convinced that I already hold most of the important information on the group and their planning, and I wonder if there is really additional information that we still need. Well, I mean, I already possess most of the most important information, so let me just share that information first.” The friendly approach tactic, the presentation of the known information, and the interruption of the phone call were implemented in the exact same manner as in the just start condition.

Post-interview questionnaire. Five minutes after the phone call was interrupted the experimenter came into the room and explained that the role-playing part was over. The participants were then asked to fill out a questionnaire concerning their perceptions of the interviewer's knowledge. The first relevant question sought to obtain a global perception of the interviewer's knowledge (“To what extent did you perceive that your police contact already had information on the group and their activities?”), with a scale ranging from 1 (to a very low extent) to 7 (to a very high extent). The next two questions regarded how actively the participants searched for information that was known to the interviewer. The participants were asked to write down all the information that the interviewer already knew (“What information on the group and their activities did your police contact already hold?”), and the final critical question was, “To what extent did you think during the phone call about what further information your police contact could have?” The scale ranged from 1 (to a very low extent) to 7 (to a very high extent). In addition, the participants were asked how motivated they felt to carry out their “task” as a source (1 = not at all motivated to 7 = very motivated), how difficult it was to understand the instructions of the study (1 = not at all difficult to 7 = very difficult), and how difficult it was to play the role of an informant (1 = not at all difficult to 7 = very difficult). The participants were then debriefed and discharged.

Coding. The participants’ written answers about what information they perceived the interviewer already knew were coded. Since some participants listed information on a high level of specificity (e.g., “S/he knew that the attack would take place on 27^th of December”), whereas others only mentioned the theme (e.g., “S/he knew the date of the attack”), we coded for a theme being mentioned or not. That is, the total amount of listed information ranged from 0 to 14.

Inter-rater reliability. Two persons coded 30% of the written answers from the open-ended question (nine for each condition), and the inter-rater reliability was calculated on the basis of these 30% (Cohen's κ = .95). The two coders discussed the results and resolved discrepancies and the final agreed-upon scores were used in the analysis. Then, one person coded the remaining 70% of the answers.

Results

Preliminary Analyses

Overall, the participants were motivated to carry out their task as a source (M = 5.30, SD = 1.45), did not find it difficult to understand the received instructions (M = 1.82, SD = 1.07), and found it moderately difficult to take the role of a source (M = 4.40, SD = 1.68). No differences were found between the two conditions with respect to the participants’ motivation, t(58) = -0.708, p = .482, d = 0.18, how difficult it was to understand the instructions, t(58) = 0.846, p = .401, d = 0.22, or how difficult it was to take the role of an informant, t(58) = -0.612, p = .543, d = 0.15.

Perception of the Interviewer's Knowledge

Known information. No difference was found between the just start condition (M = 5.33, SD = 1.67) and the traditional condition (M = 5.60, SD = 1.13) with respect to the written answers to the open-ended question regarding which information the participants believed the interviewer already knew, t(58) = -0.724, p = .236, one-tailed, d = 0.19. Thus, no support was found for Hypothesis 1.

Knowledge search. We found no difference between the just start condition (M = 4.87, SD = 1.72) and the traditional condition (M = 4.30, SD = 1.97) with respect to the extent to which participants thought about further information that the interviewer could have, t(58) = 1.189, p = .120, one-tailed, d = 0.31. Thus, Hypothesis 2 was not supported.

Global perception. As predicted in Hypothesis 3, the participants in the just start condition (M = 5.47, SD = 0.94) perceived the interviewer to hold significantly more information compared to the participants in the traditional condition (M = 4.77, SD = 1.07), t(58) = 2.692, p = .005, one-tailed, d = 0.70.

Discussion

The results offered mixed support for our expectations. As predicted, the participants in the just start condition perceived the interviewer globally to have had more information compared to the participants in the traditional condition. However, no differences were found with respect to how actively the sources searched the interviewer's knowledge for known information (i.e., the listed information that was believed to be possessed by the interviewer, and the extent to which they thought about which further information the interviewer could hold). Before discussing the results in detail we examine how sources perceive and test an interviewer's knowledge in terms of unknown information.

Experiment 2

In this experiment we explored the participants’ perception of the interviewer's knowledge gaps. We predicted that the participants in the just start condition would list comparatively less information that was unknown to the interviewer (Hypothesis 1), and that they, to a comparatively lesser extent, would search for gaps in the interviewer's knowledge (Hypothesis 2). Furthermore, we expected that the participants in the just start condition would globally believe the interviewer to have had comparatively fewer knowledge gaps (Hypothesis 3).

Method

Participants

The study included 60 participants (45 female and 15 male; 45 psychology students, 14 other students, and one employee) with a mean age of 21.73 (SD = 3.98, ranging from 17 to 37). They were randomly assigned to one of the two interview conditions (30 participants for each condition), and received course credit or 10 Euros as compensation for their participation. Originally, 62 participants were recruited, but two persons were excluded from the study (one did not fulfill the memory test criteria, and one misunderstood the instructions).

Materials and Procedure

The materials and the general procedure (background and planning, and conducting of the interviews) for Experiment 2 were exactly the same as those in Experiment 1. The two experiments differed only with respect to the post-interview questionnaire.

Post-interview Questionnaire

After the phone call was interrupted for 5 minutes, the experimenter asked the participants to fill out a questionnaire concerning their perception of the interviewer's knowledge gaps. The first critical question was about the participants’ global perception of the interviewer's knowledge gaps (“To what extent did you perceive that your police contact had knowledge gaps in terms of the group and their activities?”); the scale ranged from 1 (to a very low extent) to 7 (to a very high extent). The next two questions related to how actively the participants searched for unknown information in the interviewer's knowledge. They were asked to write down the information that they perceived the interviewer did not know (“Where did your police contact have knowledge gaps in terms of the group and their activities?”), and they rated the extent to which they searched for gaps in the interviewer's knowledge (“To what extent did you search for gaps in the knowledge of your police contact during the phone call?”) on a response scale ranging from 1 (to a very low extent) to 7 (to a very high extent). As in Experiment 1, participants were then asked how motivated they felt to carry out their “task” as a source (1 = not at all motivated to 7 = very motivated), how difficult it was to understand the study's instructions (1 = not at all difficult to 7 = very difficult), and how difficult it was to play the role of an informant (1 = not at all difficult to 7 = very difficult).

Coding. The coding of the written answers was very similar to that in Experiment 1. The only difference was that the participants wrote down information that was unknown to the interviewer (i.e., listed knowledge gaps). As in Experiment 1, we coded answers if a theme was mentioned with no indication of its level of specificity (total range: 0 to 14 themes).

Inter-rater reliability. Two persons coded 30% of the written answers (9 for each condition), and these 30% were assessed for inter-rater reliability (Cohen's κ = .92). After discussing and resolving discrepancies, the final agreed-upon scores were used in the analysis, and one person coded the remaining 70% of the answers.

Results

Preliminary Analyses

The participants were motivated to perform their role as a source (M = 5.60, SD = 1.41), did not find it difficult to understand the instructions of the study (M = 1.70, SD = .79), and found it moderately difficult to take the role of a source (M = 3.82, SD = 1.90). No differences were found between the two conditions with respect to the participants’ motivation, t(58) = 0.548, p = .586, d = 0.14, how difficult it was to understand the study's instructions, t(58) = 0.652, p = .517, d = 0.18, or how difficult it was to take the role of a source, t(58) = -0.745, p = .460, d = 0.19.

Perception of the Interviewer's Knowledge Gaps

Unknown information. As predicted in Hypothesis 1, the participants in the just start condition (M = 6.13, SD = 3.00) wrote down significantly less information that they perceived as unknown to the interviewer compared to the participants in the traditional condition (M = 7.70, SD = 3.06), t(58) = -2.000, p = .025, one-tailed, d = 0.52.

Knowledge gaps search. In support of Hypothesis 2, the participants in the just start condition (M = 4.70, SD = 1.75) searched to a significantly lesser extent for gaps in the interviewer's knowledge compared to the participants in the traditional condition (M = 5.63, SD = 1.59), t(58) = -2.168, p = .017, one-tailed, d = 0.56.

Global perception. No difference was found between the just start condition (M = 3.20, SD = 1.42) and the traditional condition (M = 3.20, SD = 1.38) with respect to the participants’ global perception of gaps in the interviewer's knowledge, t(58) = 0.000, p = 1.00, one-tailed, d = 0.00. This did not align with Hypothesis 3.

Discussion

The results confirmed two of our three predictions. As expected, we found that the participants in the just start condition listed comparatively less information that was unknown to the interviewer, and searched to a comparatively lesser extent for gaps in the interviewer's knowledge. That means, the sources in the just start condition searched comparatively less actively for gaps in the interviewer's knowledge information. However, no difference was found between the two conditions with respect to the sources’ global perception of gaps in the interviewer's knowledge. Below we discuss the results of Experiment 1 and 2 in an integrated manner.

General Discussion

The present research examined two ways of introducing the Scharff-technique's illusion of knowing it all tactic. For the traditional condition, the interviewer began presenting known information with a statement on the amount and relevance of known information. In contrast, for the just start condition, the interviewer started presenting information without such a statement. In two separate experiments we compared these two conditions and mapped how sources tested and perceived the interviewer's knowledge in terms of known (Experiment 1) and unknown information (Experiment 2).

In Experiment 1, no difference was found between the two conditions with respect to how actively the sources tested the interviewer's knowledge. However, the sources in the just start condition perceived that the interviewer possessed a larger amount of knowledge compared to the sources in the traditional condition. In Experiment 2, we found that the sources in the just start condition tested the interviewer's knowledge less actively for unknown information compared to the traditional condition. However, the two conditions did not differ in terms of the sources’ global perception of gaps in the interviewer's knowledge. In sum, the sources in the just start condition tested comparatively less actively for gaps in the interviewer's knowledge, and inferred that the interviewer held comparatively more knowledge.

The results confirmed some but not all of our predictions. Based on the framework for social hypothesis (Trope & Liberman, 1996), we expected to find a link between (a) how actively the sources tested the interviewer's knowledge in terms of known/unknown information and (b) how they–on a global level–perceived the interviewer's knowledge/knowledge gaps. Unexpectedly, we did not find this link. However, we found that in Experiment 1, the sources in the just start condition perceived the interviewer to have had a greater extent of knowledge without searching more actively for known information (vs. the traditional condition). This suggests that the hypothesis could have been confirmed without searching more actively for known information and a hypothesis confirmation strategy could have therefore been used accordingly. In Experiment 2, the sources in the traditional condition might have used a positive test strategy as they searched more actively for unknown information, but did not perceive the interviewer to have more knowledge gaps (vs. the just start condition). However, further research is needed to examine these hypothesis test strategies. Overall, in the present experiments, the just start condition outperformed the traditional condition in terms of maximizing the sources’ perceptions of the interviewer's knowledge and minimizing their focus on information that was unknown to the interviewer. The relevance of this becomes particular obvious when linking these findings to the sources’ counter-interrogation strategies (a) to reveal only information that is already known to the interviewer, and (b) to refrain from filling in crucial gaps in the interviewer's knowledge (Soufan, 2011; Toliver, 1997).

The present study came with a number of limitations as well as some fruitful input with respect to directions for future research. A rather general limitation of the illusion of knowing it all tactic is that in order to implement it, prior information is required. Furthermore, in certain situations presenting information to a source poses a risk (e.g., if a source is not credible). That is, the illusion of knowing it all tactic should only be implemented in certain situations. A limitation concerning the current study was that it was a role-play based on a student sample. However, we believe that real-life sources, who are in an information management dilemma similar to the one examined in the present study, would be comparatively more motivated to explore the interviewer's knowledge. This could mean that the just start condition may be even more effective in real-life situations. Furthermore, we decided to conduct the interviews via phone, to prohibit the participants from revealing any case-specific information, and to give them no chance to ask any questions, as this could have influenced our manipulations. However, we view the current research as a first step in examining sources’ perceptions of the illusion of knowing it all tactic and future studies should address these issues. For future studies it might be also worthwhile to vary the introductory statement (e.g., “to have some important information”) and the specificity and amount of presented information. Also, we believe that there is room for future research on how to elaborate the presented information presentation. For example, the interviewer could add in ambiguous information, as several studies have shown that such information is likely to be interpreted in favor of the hypothesis (Darley & Gross, 1983; Regan, Straus, & Fazio, 1974; Schulz-Hardt & Köhnken, 2000). Finally, it might also be worthwhile to examine how to introduce evidence in police interviews.

Conclusions

The present study examined two versions of how to introduce the illusion of knowing it all tactic. This tactic specifies that an interviewer presents already known information and is at the very core of the Scharff-technique. In conclusion, we have demonstrated that sources who were not presented with an explicit introductory statement on the presented information (a) searched less actively for information unknown to the interviewer, and (b) perceived the interviewer to have held a larger amount of prior knowledge, compared to sources who were presented with an interviewer's statement implying that s/he already had most of the important information. Further research is needed to examine more closely how to introduce and present known information in order to inflate sources’ perceptions of how much knowledge the interviewer holds about an event to be discussed.

Conflict of Interest

The authors of this article declare no conflict of interest.