EXPERIMENT 1
Participants engaged in a computer-simulated tour of an apartment, observing various objects After viewing the video, they responded with "yes" or "no" to statements about the features of the objects they had seen Following a 20-minute unrelated filler task, participants completed a final test to assess their recognition of the objects.
The experiment investigated the replicability of the negation effect using a single-item presentation list-learning paradigm Participants were shown images of simple objects and asked to respond “yes” or “no” to statements about the objects' features To replicate the encoding experience from the Mayo paradigm, subjects in Experiment 1 viewed a series of objects sequentially and were subsequently tested on the feature statements After a 20-minute filler period, a final test was conducted, which involved either object recognition (Exp 1A) or free recall (Exp 1B).
In Experiment 1, a total of 84 participants, comprising 42.9% males, were involved for partial course credit, with a mean age of 18.86 years (SD = 2.28) Within this experiment, 49 subjects participated in Experiment 1A, completing a final recognition test, while 35 subjects took part in Experiment 1B, which involved a final free recall test Notably, three subjects from Experiment 1B were excluded from the analysis due to non-compliance with instructions.
Materials and Design Stimuli included 32 images of simple objects retrieved from the Massive Visual Memory Stimuli dataset (Brady, Konkle, Alvarez, & Oliva,
In a study conducted in 2013, object images were carefully chosen to represent different attributes, such as color, and states, like open or closed The feature descriptions for each object were delivered through pre-recorded audio files featuring a female voice, with each recording lasting between 2500 to 3000 milliseconds.
Statements that were presented to participants can be found in Appendix A
This study utilized a within-subjects design to examine responses to feature questions during an initial memory test, categorized as "yes" or "no." Object memory was evaluated through performance on a subsequent final memory test In Experiment 1A, participants identified whether an object was "Present" or "Not Present" from the study phase during the final recognition test.
Experiment 1B, subjects freely recalled all of the items that they could remember seeing in the study phase
The study was conducted in three distinct phases: a study phase, an initial test phase, and a final test phase, as illustrated in Figure 1 Before the first phase commenced, participants provided informed consent and received detailed instructions regarding the experiment They were informed that they would study a set of simple objects and would later be questioned about their features, but they were not made aware that a final test would follow.
The study found that the negation effect was not significant in a pre-encoding condition, as indicated by the recognition (d = 02) and recall (d = 45) results This paper focuses on the post-encoding manipulation of negation, following the methodology established by Mayo et al (2014), while subjects engaged with the images presented through E-Prime 2.0 TM.
Participants were randomly assigned to conditions before the experiment commenced During the study phase, they viewed a complete set of 32 objects for 250 milliseconds each, separated by a 1000-millisecond inter-stimulus interval (ISI) After all object images were presented, participants engaged in an initial test phase featuring feature statements that described either correct or incorrect attributes of the studied objects These statements were randomly categorized into “yes” or “no” responses and counterbalanced to ensure each object was equally associated with both response types across all conditions.
After the study phase, participants engaged in a 20-minute filler task that involved locating sequences of numbers, akin to a word search Subsequently, they took part in a final object memory test, which in Experiment 1A, was a recognition test Participants were presented with 64 object labels, half of which corresponded to the 32 objects studied earlier, requiring them to indicate whether they recognized each item as present or not The test items were presented in a random order for each participant Additionally, subjects provided confidence ratings on a scale from 1 to 5, reflecting their certainty about their responses, and made phenomenological memory judgments by categorizing their recollections as "remember," "know," or "guess." Specific instructions for these judgments were detailed in Appendix B, and the order of confidence and memory basis ratings was counterbalanced among participants.
In Experiment 1B, the final test utilized a free recall method where participants had unlimited time to list objects they remembered from the initial phase They were prompted twice to recall all studied items, and their responses were evaluated by a research assistant who was unaware of the images used Responses were deemed correct if the assistant could identify the object, even if the terminology was not exact; for example, both "marker" and "highlighter" would be accepted for "highlighter." After the memory test, participants were debriefed and released from the study.
In this study, I aim to replicate the negation-induced forgetting effect observed in Mayo et al (2014) by reporting Bayes Factors (BF10) alongside traditional null hypothesis significance tests (NHST) Utilizing Bayes factors allows for a robust comparison of evidence strength between two models: one that posits no significant negation effect, representing the null hypothesis.
3 Subjects provided these responses for both “old” and “new” responses In the results, I focus on the phenomenological memory judgments for only correct “old” responses
Model 0), and one in which there is a significant effect (the alternative hypothesis, Model
The BF10 factor quantifies the probability of data under the alternative hypothesis (Model 1) compared to the null hypothesis (Model 0), with values below 1 being inverted for clarity For subject-level effects, I will utilize the effect size from Mayo et al (2014) Experiment 1, reported as d = 0.53 [0.19, 0.87], as the prior distribution In item-level analyses, a standard Cauchy prior of 0.707 will be applied to compute BF10 The interpretation of BF10's magnitude indicates the strength of evidence: values between 0 to 3 signify anecdotal evidence, 3 to 10 denote substantial evidence, 10 to 100 reflect strong evidence, and values exceeding 100 represent decisive evidence (Jeffreys, 1961).
The initial memory test aimed to gather "yes" and "no" responses from participants regarding object features, consisting of 32 statements—half accurate and half inaccurate A paired-samples t-test evaluated how these responses influenced the accuracy of feature identification Results showed that participants were similarly accurate for both "yes" responses (M = 72, SE = 02) and "no" responses (M = 70, SE = 02), with no significant difference found (t(48) = 1.52, p = 14, d = 0.17 [-0.05, 0.40]).
The final memory test focused on two key outcome measures: the proportion of conditionalized errors and the phenomenological memory bases for accurately recognized objects Conditionalized errors are defined as items that received a correct response during the initial memory test but were incorrectly identified as “Not Present” in the final test Additionally, proportions of Remember and Know responses were calculated for objects accurately recognized from the first part of the experiment, with Know responses adjusted for independence by dividing the number of Know responses by the opportunities to respond Know.
A paired-samples t-test was conducted to evaluate how initial memory test responses (yes or no) impacted the proportion of errors in the final memory test Specifically, this analysis focused on studied items that were correctly answered in the initial test but later failed to be recognized The results indicated that participants made more errors after responding "no" (M = 10, SE = 02) compared to those who responded "yes" (M = 07).
SE = 01), a statistically significant negation effect was not observed, t(48) = 1.49, p .14, d = 27 [-.09, 62]
EXPERIMENT 2
The sampling plan was designed through a power analysis aimed at identifying a small to medium effect size Utilizing G*Power 3.1 (Faul et al., 2009), the calculated sample size was n = 68, which would provide a statistical power of 0.90 for a two-tailed dependent t-test, ensuring robust results for the study.
In total, the pre-registered laboratory sample (Experiment 2a) was comprised of
A total of 75 students from Iowa State University participated in this experiment for partial course credit, with an average age of 19.54 years (SD = 1.05), and 45.3% of the participants were male Due to a program error, demographic information for one participant was not available Additionally, data from ten subjects were excluded from the analysis: nine were non-native English speakers, and one was identified as a multivariate outlier using Mahalanobis distance (Tabachnick & Fidell, 1996).
In Experiment 2b, a total of 80 participants (41.40% male) were recruited through Amazon's Mechanical Turk, with a mean age of 35.49 years (SD = 10.49) Each worker received $1.00 for their participation in the task It is important to note that data from ten participants were excluded from the analysis due to incorrect responses to attention checks.
Materials This experiment was presented to subjects in the laboratory using E-
The study utilized Prime 2.0 TM and Mechanical Turk participants through Qualtrics, featuring an 8-minute, 15-second video created by the primary author of the original experiment (Mayo et al., 2014) This video showcased a digitally simulated apartment tour, highlighting various everyday household objects Detailed instructions and test items from Experiment 2 are available in Appendix C and Appendix D To ensure reliability, four different randomizations of the initial test were programmed and counterbalanced among participants.
The study utilized a fully within-subjects design, focusing on yes-no responses to statements during an initial memory test The online version of the study featured only minor procedural variations compared to the original format.
The original study excluded participants who scored below 50% on initial memory tests; however, this exclusion was not reflected in the registration I analyzed the data both with and without these six subjects, and the conclusions remained consistent To adhere to the pre-registration, the results here include all subjects, regardless of their initial accuracy Informed consent was obtained from participants, who were instructed that they would watch a brief film and answer related questions Additionally, online participants completed attention check questions before the experiment to ensure engagement, while the overall procedure remained consistent between laboratory and online samples, aside from extra attention checks after a filler task.
Participants were instructed to view a stimulus video before completing a short memory task involving 16 feature statements about objects in an apartment These statements included both congruent features, which were answered with "yes," and incongruent features, answered with "no." The statements were randomly divided into four conditions to ensure counterbalancing of correct responses Following the memory task, participants engaged in a 20-minute unrelated filler task that involved constructing words from 15-character prompts.
Participants were tasked with creating new words from a set of base words, such as "random" and "mentors," using only the letters provided They utilized a computer program to generate as many words as possible within a four-minute timeframe for each of the five base words.
Once all five base words had been completed, subjects were administered a final object recognition test in which they were presented with a series of object labels (e.g.,
Participants were shown a video featuring various objects, including a chair, and later asked to identify whether these items appeared in the video The final test comprised 32 objects: 16 that were included in the initial memory test and 16 new objects not present in the video This selection mirrored the methodology used by Mayo and colleagues (2014) After the final assessment, participants were debriefed and released from the study.
The initial memory test involved 16 statements about object features from an apartment video, with half requiring "yes" responses for correct features and the other half needing "no" for incorrect features Accuracy was measured by the ratio of correct responses, revealing that participants were significantly more accurate in providing "no" responses (M = 75, SE = 02) compared to "yes" responses (M = 65, SE = 03), with a t-value of 3.07 (p = 003, d = 0.53 [.18, 88]) This finding contrasts with the results of Mayo et al (2014), where no accuracy difference was observed between "yes" and "no" responses in the initial test.
The final memory test evaluated participants' recognition of 32 object labels, which included 16 objects featured in a video and 16 novel objects not shown Participants were tasked with identifying whether each object was present in the apartment they viewed Accuracy was measured by the proportion of correctly identified items, with an overall mean accuracy of 85% (M = 85, SD = 19), indicating that subjects performed well on the test.
The primary focus of the study was the errors observed in the final memory test, specifically regarding objects that were present in the video but incorrectly identified as “not present.” This analysis was limited to items that participants accurately identified in the initial memory test The results indicated a significant negation effect, where objects linked to “no” responses (M = 14, SE = 02) resulted in more memory errors compared to those linked to “yes” responses (M = 07, SE = 02), with a statistically significant difference (t(64) = 3.08, p = 003, d = 0.50) The strong Bayes Factor (BF10 = 11.16) supports the presence of a negation effect at the subject level Additionally, an item-level analysis revealed a marginally significant negation effect (t(15) = 2.01, p = 06, d = 60), with a BF10 of 1.26 indicating anecdotal evidence for this effect among the 16 target items.
Initial Memory Test With regard to the initial memory test involve yes-no responses, accuracy was again greater for statements requiring a “no” response (M = 66,
SE = 02) than for statements requiring a “yes” response (M = 56, SE = 03), t(69) = 2.51,
The registration failed to include item-level analysis, which I now present to align with the findings of Mayo et al (2014), showing a significant effect (p = 014, d = 0.46 [.09, 82]) While this effect differs from Mayo et al (2014), it successfully replicates the laboratory sample results observed in Experiment 2A.
The final memory test assessed performance through conditionalized errors, revealing a significant negation effect Objects linked to "no" responses (M = 16, SE = 02) resulted in more memory errors compared to those linked to "yes" responses (M = 08, SE = 02), with statistical significance (t(69) = 2.54, p = 014, d = 0.41) An estimated BF10 of 3.04 indicates strong evidence for the negation effect at the subject level Further analysis at the item level confirmed this effect across 16 target items, showing that objects associated with "no" responses were more often miscategorized as "Not Present" (M = 16, SE = 03) than those associated with "yes" (M = 07, SE = 02), with significant results (t(15) = 2.88, p = 01, d = 88) and an estimated BF10 of 4.90, supporting the presence of the negation effect.
Results – Combined Samples Analysis (Experiments 2a and 2b)
Both laboratory (Experiment 2a) and online (Experiment 2b) samples exhibited similar results on the final test, as indicated by a 2 x 2 mixed model ANOVA evaluating the effects of sample type (laboratory, online) and responses to feature statements (yes, no) on conditionalized final memory test errors The analysis revealed no significant main effect of sample (F(1, 133) = 23, p = 63) or interaction between yes-no responses and sample (F(1, 133) = 01, p = 94) However, aligning with the findings of Mayo et al (2014), a significant negation effect was identified, where target objects linked to “no” responses (M = 15, SE = 02) resulted in more memory errors compared to those associated with “yes” responses (M = 08, SE = 01), F(1, 133).
EXPERIMENT 3
After corresponding with the primary author and closely examining the procedures and materials used in Mayo et al (2014), I hypothesized that the discrepancies between the original experiment and its conceptual replication may stem from differences in stimulus presentation The original study showcased a significant negation effect through a video tour of an apartment, which featured numerous household objects and furniture Notably, 16 specific items were selected for testing post-video presentation, while I estimated that each tested object was accompanied by 3 to 4 untested items within the apartment.
In Experiment 1, the replication efforts followed the foundational principles established by Mayo et al (2014), involving a sequence where subjects were presented with several objects, underwent an initial memory test, engaged in a 20-minute filler task, and concluded with a final memory assessment.
The paradigm used in Experiment 1 lacked additional non-tested items during the study task, unlike the Mayo et al (2014) study, which may have increased memory load for participants This suggests that the task in Mayo et al (2014) was more challenging, as encoding had to be distributed across both target and non-tested items Research in misinformation indicates that individuals are more vulnerable to false information when their memory quality is compromised, influenced by factors such as time elapsed (Loftus, Miller & Burns, 1978) or the centrality of information at the time of encoding (Wilford, Chan, & ).
Research by Mayo et al (2014) suggests that distributing encoding over multiple items can lead to reduced memory strength In Experiment 3, the number of items encoded was varied to assess whether a higher memory load would amplify the negation effect, indicating that increased cognitive demands may weaken memory retention among participants.
In exploring the moderating role of alternatives in the negation effect, research indicates that the context of a yes-no question—whether it suggests two or multiple alternatives—significantly influences attention and memory Specifically, when multiple alternatives are presented, individuals tend to focus their attention in a way that enhances the negation effect Previous studies have shown that negated descriptions linked to less accessible opposing constructs are more prone to memory errors compared to those with readily available opposites In Experiment 3, I manipulated the number of alternatives in yes-no questions, contrasting feature statements with clear opposites (e.g., "not open" implies "closed") against those associated with multiple alternatives (e.g., "not red" suggests "blue," "green," etc.) Given the findings that negations tied to multiple options can paradoxically heighten attention and lead to increased meaning-based memory errors, I hypothesized that such negations would yield more errors in the final memory test This factor has yet to be examined within an object recognition framework, limiting the development of a more nuanced theoretical prediction.
A total of 190 students from Iowa State University participated in this study, with an average age of 19.35 years (SD = 1.05), and 46.3% of the participants were male The planned sample size was calculated to be n = 136 using G*Power 3.1 to achieve a power of 0.90 for a repeated measures ANOVA with a small to medium effect size However, demographic information was missing for two subjects due to a program error, and eleven participants were excluded from the analysis for being non-native English speakers (N = 8) or color-blind (N = 3) Consequently, the final analysis included unequal between-subject conditions, with the load-present condition consisting of 88 participants and the load-absent condition comprising 91 participants.
In this experiment, the number of stimuli was increased to 48 images sourced from the Massive Visual Memory Stimuli dataset (Brady et al., 2013) The selected object images were varied based on two key features: emptiness and openness, along with multi-option features such as type, color, and shape Feature statements were developed for the objects under study.
I conducted analyses both including and excluding participants who scored below 50% on the initial memory test (N = 6), and the findings remained consistent The statements were delivered through pre-recorded audio files featuring a female speaker, with each recording lasting between 2500ms and 3500ms.
A 2 x 2 x 2 mixed design was used to assess the influence of memory load
In this study, object memory was evaluated by manipulating memory load between subjects, with a 4:1 ratio of distractor to target objects based on Mayo et al (2014) Participants were divided into two groups: one group encoded only the 48 target objects (load-absent condition), while the other group studied the same 48 target objects along with 144 additional distractors (load-present condition) The number of possible alternatives and responses to feature statements were manipulated within subjects Object memory performance was measured through a final recognition test, where participants indicated whether an object was "Present" or "Not Present" from the study phase.
In the study, participants faced four distinct types of questions for each memory load group: yes, multi-option questions prompted a “yes” response without a specific opposing construct (e.g., “The canister was pink”); yes, two-option questions also elicited a “yes” response but were linked to an easily accessible opposing construct (e.g., “The mp3 player was turned on”); no, multi-option questions required a “no” response without invoking a particular opposing representation (e.g., “The street sign was rectangular”); and no, two-option questions elicited a “no” response associated with an easily accessible opposing construct (e.g., “The pencil cup was empty”) For a complete list of the questions used in Experiment 3, refer to Appendix E.
The experiment consisted of three distinct phases: a study phase, an initial test phase, and a final test phase, with a 20-minute filler task separating the initial and final tests Participants were randomly assigned to either the load-present or load-absent conditions.
Before the experiment commenced, participants gave their informed consent and were informed they would study a series of object images They then entered the study phase, where they examined either 48 objects in a load-absent condition.
In a study involving 196 objects presented for 1000 milliseconds each, participants underwent an initial memory test after viewing all items This test included 48 feature statements requiring yes/no responses, with half of the statements prompting a "yes" and the other half a "no." The feature statements were divided into two categories: multi-option constructs and two-option constructs, ensuring a balanced presentation of response types for each object throughout the experiment Following the test, participants completed a word-construction filler task similar to that used in Experiment 2 (cf Mayo et al.).
The final test phase followed the 20-min filler task period Subjects completed an object recognition test comprised of 96 object labels that they were to categorize as
“Present” or “Not Present” in the study phase Forty-eight of the object labels
In Experiment 1, seven different encoding times were tested, ranging from 5000 ms to 250 ms While variations in encoding time did not significantly impact initial test accuracy (ranging from 83% to 92%), they often resulted in ceiling effects for final test accuracy (87% to 93%), which may have masked the detection of a negation effect The final test included 48 target objects from the study phase, categorized as "Present" under both load-absent and load-present conditions, alongside 48 novel filler objects categorized as "Not Present." Participants were asked to express their confidence in each judgment on a scale from 50% to 100% in 10% increments Additionally, for objects identified as "Present," subjects indicated their recognition basis—whether it was based on memory, familiarity, or guessing.
“Remember” vs a “Know” judgment These instructions were adapted from the instructions used in Meissner, Brigham, and Butz (2005) and can be found in Appendix
F When subjects completed the final recognition test, they were debriefed and dismissed from the study
The initial memory test included 48 statements related to the features of objects presented during the study phase Of these statements, half accurately described the objects, prompting "yes" responses, while the other half contained incorrect descriptions that required "no" answers.
GENERAL DISCUSSION
Research on language comprehension highlights the differences in how negations, such as "no," are understood compared to affirmatives like "yes." While numerous studies have investigated the impact of negation on attention, the effects on memory for objects have only recently been explored This article aims to fill the gap in existing research by replicating previous findings and evaluating potential moderators of the negation effect in object recognition.
Replicability of the Negation Effect
Research consistently demonstrates a negation effect in memory, revealing that items linked to a "no" response during an initial test are less likely to be recalled in a subsequent test compared to those associated with a "yes" response Experiment 1 employed a list-learning paradigm to effectively replicate this finding.
The "negation-induced forgetting" effect, as examined in Mayo et al (2014), showed a non-significant memory impairment for objects linked to "no" responses, although the effect size was close to the original study's confidence interval To replicate these findings, Experiment 2 followed established replication best practices, including correspondence with the primary author for procedural alignment and material acquisition, a power analysis for robust statistical power, and pre-registration with the Open Science Framework This replication revealed significant negation effects in both laboratory and online samples, with the online sample showing a smaller effect size than the original study Experiment 3 explored two moderators of the negation effect: the number of alternatives related to a construct and memory load during encoding While memory load did not impact the negation effect's magnitude, constructs with multiple options yielded a greater negation effect compared to those with two options.
A random effects model was employed to conduct an average weighted effect size analysis across five studies (Exp 1A, 1B, 2A, 2B, and 3), revealing a small to medium effect size (k = 5, N = 395, d = 0.37 [0.28, 0.47]) Furthermore, a meta-analytic Bayes factor (BF10) analysis indicated a value significantly greater than 100, providing strong evidence for the existence of a negation effect, thereby rejecting the null hypothesis of no effect This aligns with the original effect size of d = 0.53 reported by Mayo et al.
The 2014 study was excluded from the average weighted effect size, as shown in Figure 5 This phenomenon is not uncommon in replication studies, as subsequent replications often yield weaker evidence compared to the initial reported effect size, as noted by the Open Science Collaboration in 2015.
Theoretical Mechanisms Leading to Negation
The current studies aim to investigate the replicability of the negation effect while also exploring the mechanisms behind memory impairment after a "no" response Despite this focus, the reasons for memory deficits associated with negation remain unclear Mayo et al (2014) suggested that these impairments may differ from the inhibition of competing concepts, such as retrieval-induced forgetting, as proposed by Anderson and Bjork.
According to Bjork (1994), negating a feature of an item inhibits the representation of the entire item In the context of retrieval-induced forgetting, this inhibition affects non-tested competitive material Mayo et al (2014) propose a spreading inhibition mechanism that focuses on the inhibition of the tested material, or the object itself This mechanism posits that negating a statement about an incorrect feature leads to the creation of a temporary mental representation of the object with that incorrect feature Consequently, this transient representation is completely suppressed, inhibiting the representation of the object with the correct feature during subsequent tests.
The "spreading inhibition" mechanism operates through attribute-object relationships highlighted in tested statements It suggests that negation leads to the inhibition of an object by affecting its associated attributes or features However, this mechanism does not explain variations in the strength of the negation effect depending on the number of alternatives related to a tested feature.
In Experiment 3, a significant negation effect was noted when participants were presented with feature statements implying multiple alternatives, as opposed to just two This suggests that when subjects considered a variety of potential attribute-object relationships, their recollection of the true relationship diminished, leading to increased uncertainty Conversely, two-option feature statements provided a clearer context, enhancing memory recall by framing the situation as “if not this object, then that object.” The ambiguity of multi-option statements, however, does not support this clear contextualization, resulting in less effective memory retrieval.
The difference between two-option and multi-option feature statements in memory tests can be likened to true/false versus multiple-choice questions Research by Brown, Schilling, and Hockensmith (1999) found a negative suggestion effect, where exposure to incorrect alternatives led to decreased performance on memory tests, particularly with multiple-choice formats compared to cued-recall tests In their study, participants who completed an initial cued-recall test on trivia facts showed improved performance on a subsequent memory test when not exposed to incorrect alternatives Conversely, Roediger and Marsh (2005) demonstrated that a higher number of alternatives in an initial multiple-choice test resulted in more incorrect responses on a later cued-recall test Both studies indicate that exposure to incorrect options impairs final memory performance, suggesting that considering multiple alternatives may reduce confidence in correct responses and destabilize memory traces.
The attentional centrality of an item may serve as an unexplored moderator in the studies conducted In the conceptual replications (Exp 1 and Exp 3), items were centrally framed during encoding, with participants studying one item at a time in the center of the screen In contrast, the original Mayo et al (2014) Experiment 1 and the current Exp 2 utilized a video format where the centrality and attention to items varied Participants in the Mayo et al paradigm viewed videos with multiple items in each frame, allowing them to choose where to focus their attention While participants in the list-learning paradigm were encouraged to encode all presented items, those in the Mayo et al experiment had greater control over their attention allocation Despite using central framing, a small negation effect was observed, which was less pronounced than in previous findings Research in misinformation suggests that individuals are more vulnerable to inaccuracies regarding peripheral details than central ones.
The distinction between central and peripheral processing may stem from inadequate encoding, which can hinder monitoring when misinformation is encountered Consequently, accurately responding "no" could negatively impact later memory more significantly for peripheral items than for central ones Although errors increased in Experiment 3 with higher memory load, the effect of degraded encoding did not affect the strength of the negation effect Therefore, it remains uncertain whether altering centrality will influence the magnitude of the negation effect or simply elevate overall error rates.
Practical Implications of the Negation Effect in Memory
Closed-ended questions, such as yes/no or true/false inquiries, are common in everyday communication Recent experiments have focused on how responses to these questions can impact memory recall regarding an entity or event The findings indicate that providing a correct "yes" or "no" answer can negatively affect what individuals remember, highlighting the potential detrimental effects of such responses on memory retention.
After a crime, police investigators often interview witnesses to gather detailed accounts of the event These interviews typically involve specific questioning to elicit information (Snook et al., 2012) Interestingly, research indicates that a witness's accurate response, such as saying "no" to whether the perpetrator wore a blue baseball cap, may negatively impact their overall memory of the perpetrator wearing any cap at all.
Recent studies provide compelling evidence for a negation effect in memory, indicating that the number of alternatives related to a tested concept significantly influences the extent of memory impairment This effect arises because the absence of a readily accessible opposite representation heightens vulnerability to the negation effect, potentially due to the inhibition of the object's representation or reduced recollection based on the type of feature tested Future research should explore the theoretical mechanisms behind this effect and identify potential moderators In the context of investigations, it is crucial for investigators to recognize that certain questioning techniques can enhance the accuracy of eyewitness accounts, while others, such as specific questioning, may contribute to memory deficits.
Figure 5 Forest plot of negation effect sizes for Mayo et al (2014) Experiment 1 and the present thesis studies
Anderson, M C., Bjork, R A., & Bjork, E L (1994) Remembering can cause forgetting: retrieval dynamics in long-term memory Journal of Experimental
Psychology: Learning, Memory, and Cognition, 20(5), 1063-1087
Brady, T F., Konkle, T., Alvarez, G A., & Oliva, A (2013) Real-world objects are not represented as bound units: independent forgetting of different object details from visual memory Journal of Experimental Psychology: General, 142(3), 791
Brandt, M J., IJzerman, H., Dijksterhuis, A., Farach, F J., Geller, J., Giner-Sorolla, R.,
& Van't Veer, A (2014) The replication recipe: What makes for a convincing replication? Journal of Experimental Social Psychology, 50, 217-224
Brown, A S., Schilling, H E., & Hockensmith, M L (1999) The negative suggestion effect: Pondering incorrect alternatives may be hazardous to your knowledge Journal of Educational Psychology, 91(4), 756
Camp, G., Wesstein, H., & Bruin, A B (2012) Can questioning induce dorgetting?
Retrieval-induced forgetting of eyewitness information Applied Cognitive
Chan, J C., Thomas, A K., & Bulevich, J B (2009) Recalling a witnessed event increases eyewitness suggestibility: The reversed testing effect Psychological
Craik, F I., & Tulving, E (1975) Depth of processing and the retention of words in episodic memory Journal of Experimental Psychology: General, 104(3), 268
Cumming, G (2012) Understanding the new statistics: Effect sizes, confidence intervals, and meta-analysis New York: Routledge Taylor and Francis Group
Engelkamp, J., & Hửrmann, H (1974) The effect of non-verbal information on the recall of negation The Quarterly Journal of Experimental Psychology,26(1), 98-105
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G (2009) Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses Behavior
Garven, S., Wood, J M., Malpass, R S., & Shaw III, J S (1998) More than suggestion:
The effect of interviewing techniques from the McMartin Preschool case Journal of Applied Psychology, 83(3), 347
Geiselman, R E., Fisher, R P., MacKinnon, D.P., & Holland, H L.(1986) Enhancement of eyewitness memory with the cognitive interview American Journal of
Gough, P B (1966) The verification of sentences: The effects of delay of evidence and sentence length Journal of Verbal Learning and Verbal Behavior, 5(5), 492-496
Hasson, U., & Glucksberg, S (2006) Does understanding negation entail affirmation?:
An examination of negated metaphors Journal of Pragmatics, 38(7), 1015-1032
Hope, L., Ost, J., Gabbert, F., Healey, S., & Lenton, E (2008) “With a little help from my friends…”: The role of co-witness relationship in susceptibility to misinformation Acta Psychologica, 127(2), 476-484
Kaup, B., & Zwaan, R A (2003) Effects of negation and situational presence on the accessibility of text information Journal of Experimental Psychology: Learning,
Loftus, E F (2005) Planting misinformation in the human mind: A 30-year investigation of the malleability of memory Learning & Memory, 12(4), 361-366
Loftus, E F., Miller, D G., & Burns, H J (1978) Semantic integration of verbal information into a visual memory Journal of Experimental Psychology: Human
MacDonald, M C., & Just, M A (1989) Changes in activation levels with negation Journal of Experimental Psychology: Learning, Memory, and
Maciuszek, J (2013) Don’t pay attention to what you see! Negative commands and attention bias Polish Psychological Bulletin, 44(1), 70-84
Mason, W., & Suri, S (2012) Conducting behavioral research on Amazon’s Mechanical
Mayo, R., Schul, Y., & Burnstein, E (2004) “I am not guilty” vs “I am innocent”: successful negation may depend on the schema used for its encoding Journal of
Mayo, R., Schul, Y., & Rosenthal, M (2014) If you negate, you may forget: Negated repetitions impair memory compared with affirmative repetitions Journal of
Meissner, C A., Brigham, J C., & Butz, D A (2005) Memory for own- and other-race faces: A dual-process approach Applied Cognitive Psychology, 19(5), 545-567
Meyer, D E (1975) Long-term memory retrieval during the comprehension of affirmative and negative sentences Studies in Long-Term Memory London:
Orenes, I., Beltrán, D., & Santamaría, C (2014) How negation is understood: Evidence from the visual world paradigm Journal of Memory and Language, 74, 36-45
Raaijmakers, J G., Schrijnemakers, J M C., & Gremmen, F (1999) How to deal with
“the language-as-fixed-effect fallacy”: Common misconceptions and alternative solutions Journal of Memory and Language, 41, 416-426
Schooler, J W., & Engstler-Schooler, T Y (1990) Verbal overshadowing of visual memories: Some things are better left unsaid Cognitive Psychology, 22(1), 36-71
Snook, B., Luther, K., Quinlan, H., & Milne, R (2012) Let’em talk! A field study of police questioning practices of suspects and accused persons Criminal Justice and Behavior, 39(10), 1328-1339
Spies, J R., Nosek, B A., Miguez, S., Blohowiak, B B., Cohn, M., Bartmess, E., et al
(2012) Open Science Framework Open Science Collaboration Available online at: http://openscienceframework.org/ (accessed June 3, 2016)
Tabachnick, B G., & Fidell, L S (1996) Using multivariate statistics New York, NY:
Wason, P C (1961) Response to affirmative and negative binary statements British
Wegner, D M., Schneider, D J., Carter, S R., & White, T L (1987) Paradoxical effects of thought suppression Journal of Personality and Social Psychology, 53(1), 5-
Wilford, M M., Chan, J C., & Tuhn, S J (2014) Retrieval enhances eyewitness suggestibility to misinformation in free and cued recall Journal of Experimental
Wright, D B., & Stroud, J N (1998) Memory quality and misinformation for peripheral and central objects Legal and Criminological Psychology, 3(2), 273-286
Yonelinas, A P., & Jacoby, L L (1995) The relation between remembering and knowing as bases for recognition: Effects of size congruency Journal of Memory and Language, 34(5), 622-643
Initial Memory Test – “Yes” Initial Memory Test – “No” Object on final test
The umbrella was open The umbrella was closed umbrella The flag was upright on the mailbox
The flag was down on the mailbox mailbox