Gun Focus Effect Revisited

Abstract

This study attempts to clarify the relationship between message emotional tone and memory vividness, specifically gun focus effect. Based on findings in neuroimaging research, it is claimed here that positive and negative messages trigger different processing mechanisms in the brain and therefore later detail retrieval is dependent on the degree to which the same processing paths are activated. Overall, data support an assumption that visual recognition aids negative arousing content detail memory more than positive. In contrast, cued recall procedure is more helpful for recalling positive detail. For central detail the matching cues work together with the arousing content to boost memory specificity. Yet for peripheral detail, the matching cues help overcome the harmful effects of arousing content by reducing forgetting. Thus visual recognition minimizes gun focus effect for negative visual content and enhances it for positive. At the same time, a cued recall task minimizes gun focus effect for positive visual content and enhances it for negative.

Keywords

gun focus memory for detail encoding recall vs. recognition emotion and memory interactions

Gun focus effect has been observed in eyewitness testimonies where a witness may remember the details of an assailant’s gun but be unable to remember details of the assailant. In memory literature, this is also known as a memory narrowing phenomenon (first established by Easterbrook, 1959 and replicated numerous times) showing that the central details of highly arousing negative events are typically remembered better than those of neutral events, but memory for peripheral details is often poorer for negative events than for neutral events (Christianson & Loftus, 1991; Clifford & Hollin, 1981; Kebeck & Lohaus, 1986; Libkuman, Stabler, & Otani, 2004; Loftus & Burns, 1982).

Surprisingly little attention has been devoted to this phenomenon in communication research. However, in a modern multimedia environment, understanding how details are encoded and retrieved is essential to understanding how media is processed. This knowledge is instrumental to message producers as they compete for viewer attention.

Mediated content is usually emotionally charged. Research agrees that emotional content is preferentially processed in comparison to nonemotional (Bucy & Newhagen, 1999; Lang, 2006; Lang, Dhillon, & Dong, 1995; Shoemaker, 1996). Results, however, are contradictory when it comes to demonstrating benefits of processing positive versus negative content. Some studies report strong evidence for negative content advantages for attention and memory (e.g., Bolls, Lang, & Potter, 2001; Bradley, Angelini, & Lee, 2007; Goss, Neuliep, & O’Hair, 1985; Lang, Newhagen, & Reeves, 1996). Yet others report beneficial processing for positive content (Talarico, Berntsen, & Rubin, 2009), especially for some groups (Grabe & Kamhawi, 2006). These studies, however, focus primarily on gist memory. The existing research on memory for detail, such as contextual cues, sources, and locations, is also inconclusive on the role of valence. Although gun focus is a well-established phenomenon, some researchers observed memory broadening for negative content (Kensinger, Garoff-Eaton, Schacter, 2007). Similar inconsistencies emerge from observing positive content effects. Heuer and Reisberg (1990) reported memory narrowing for positive content, while others showed the broadening effect (Yegiyan & Lang, 2010; Yegiyan & Yonelinas, 2011).

This study is an attempt to clarify the relationship between message emotional tone and memory vividness. Based on findings in neuroimaging research, it is claimed here that positive and negative messages trigger different processing mechanisms in the brain, and therefore detail retrieval is dependent on the degree to which the same processing paths are activated. The better the match, the more specific are the memories. In practical terms, different memory tests may lead to opposing results for both positive and negative events and can lead to the elimination of the gun focus effect under certain conditions.

Message Valence Affects Both Encoding and Retrieval Strategies

Memory enhancing properties of arousal are well established. Yet when examining the specificity of the memories of arousing events, most research reports memory narrowing effects, with very detailed memory for central aspects of the events and very little memory for peripheral detail (see Buchanan & Adolphs, 2002; Mather, 2007; Reisberg & Heuer, 2004 for theoretical explanations). Memories for positive events are shown to be less focused on central aspects and more open to encoding contextual information, thus producing a memory-broadening effect. (Mickley & Kensinger, 2008; Yegiyan & Lang, 2010). This seems to suggest that negative events produce superior memory for intrinsic detail (detail related to the item itself), while positive events enhance memory for extrinsic detail (detail related to contextual aspects). More importantly, the types of the well-remembered detail also differed. Negative events’ intrinsic memories were more sensory in nature, such as those related to the color or the make of objects. Positive events’ extrinsic memories were more semantic, related to aspects of conceptual understanding, such as spatial position, temporal location, and amount (e.g., Fredrickson & Branigan, 2005; Gasper & Clore, 2002; Rowe, Hirsh, & Anderson, 2007).

Neuroimaging research suggests that valence triggers different types of activation during negative and positive message processing. Negative information leads to higher activity within sensory processing regions, including the occipital (visual) cortex and along the fusiform gyrus (a region specialized for processing high-level features of objects and faces and for encoding those stimuli; Bernstein, Berg, Siegenthaler, & Grady, 2002; Garoff, Slotnick, & Schacter, 2005; Kirchhoff, Wagner, Maril, & Stern, 2000; Kuskowski & Pardo, 1999). The encoding of positive information is tied to disproportionate recruitment of lateral prefrontal and temporal regions that often have been implicated in the processing of semantic or conceptual information (Dobbins &Wagner, 2005; Poldrack, Selco, Field, & Cohen, 1999).

Neuroimaging studies have also shown that when we retrieve emotional information we engage those regions that we initially recruited to process that event (e.g., Kahn, Davachi, & Wagner, 2004; Nyberg, Habib, McIntosh, & Tulving, 2000; Vaidya, Zhao, Desmond, & Gabrieli, 2002;Wheeler, Petersen, & Buckner, 2000). Given that retrieval is most successful when there is a match between the processes engaged during encoding and those engaged during retrieval (e.g., Craik & Lockhart, 1972), it is possible to assume that the way in which people encode information (i.e., in a perceptual vs. conceptual manner) would affect the types of retrieval processes or retrieval cues that would be most effective in guiding memory. If information was encoded using primarily sensory regions of the brain, then memory tests that enable activation of sensory regions would be most instrumental in aiding accurate recollection. Yet if encoding is based primarily on activation of brain regions responsible for conceptual processing, then memory tests that activate similar regions may be more helpful.

Recognition and Cued Recall as Facilitators of Sensory and Semantic Processing

Recognition and cued recall have been successfully used in measuring episodic memory, with recognition almost always producing better accuracy results than cued recall. This advantage has been attributed to the larger number of cues available when making recognition decisions compared to those available with cued recall response. This is not surprising, since during recognition participants are presented with an item as a whole, while cued recall provides only a description of, or a partial reference to, the item (see Shiffrin, 1999 for an in-depth review).

However, when it comes to measuring memory for emotional content detail, recognition and cued recall may provide additional advantages and disadvantages when remembering positive versus negative events. A detail recognition task usually consists of presenting only partial information: central items only or peripheral items only. For example, after viewing a scene of a car crash on a street, participants in some studies are presented with a picture of the car (central), or with a picture of the street (peripheral). In other studies, participants are exposed to the same picture of the crash but with the central item (car) masked during peripheral item recognition and the peripheral item (street) masked during central item recognition.

In addition, the nature of the task is such that not only old (previously seen) items need to be identified, but also new (not previously seen) items need to be correctly rejected. These types of decisions require that a participant compare intrinsic features of items (those related to the item itself) such as the color of the car or size of the building. This type of information is sensory in nature (in this case, visual) and thus would activate the sensory regions of the brain. Because these same regions are dominantly engaged during negative content encoding, they should aid retrieval of the items associated with negative events compared to positive. If this is the case, memory should not be diminished for negative detail at high compared to low arousing content level. However, we should see evidence of impairment for positive content, especially for peripheral detail, as has been demonstrated in earlier research. Based on this, the following hypotheses are posed:

Hypothesis 1: Recognition for negative content detail will be such that memory for both central and peripheral detail will show a tendency for improvement and not decrease for high arousing content compared to that of low.

Hypothesis 2: Recognition for positive content detail will be such that memory for detail will decrease for high arousing content compared to that of low, especially for peripheral detail.

In contrast to recognition, cued recall involves a verbal description of an event (gist) as a cue, then a question is asked about a specific detail. The task is to recall that specific detail. The response in this case is not based on a comparison of intrinsic components; rather, it is schema based. Participants activate a general idea of what the events of that sort (e.g., weddings, funerals, or crime scenes) include, and then try to match their specific memory for that particular episode with this activated schema. Although the task involves recall of sensory content, it requires a higher degree of reliance on conceptual and semantic cues and thus helps activate processes similar to those involved in positive event encoding. Based on this, the following hypotheses are posed:

Hypothesis 3: Cued recall for negative content detail will be such that memory for detail will decrease for high arousing content compared to that of low, especially for peripheral detail.

Hypothesis 4: Cued recall for positive content detail will be such that memory for both central and peripheral detail will show a tendency for improvement and not decrease for high arousing content compared to that of low.

Method

Design

This experiment had Valence (2) × Arousal (2) × Message Repetition (3) within subjects design. The valence factor had two categories: positive and negative. The arousing content factor had two levels: low and high. There were three video clips within each valence by arousing content cell, which made up the message repetition factor.

Stimulus

The stimulus consisted of 12 messages selected through a pretest from a pool of television messages and feature films. Each message averaged 60 seconds in duration. Topics across arousing content levels were kept consistent for both positive and negative emotional tone and were concerned with nature (pollution vs. gardening), food and food consumption (nausea vs. thanksgiving dinner meal), romance (break up, separation vs. date, reunion) and sexual encounters (rape vs. erotica), and sports (losing vs. winning).

Manipulation Check

Sixty-eight undergraduates indicated how the 50 preselected messages made them feel on 9-point arousal, positivity, and negativity scales. Participants viewed each clip, rated how it made them feel, and then were asked to describe in one sentence what happened in the clip they had just seen (the gist of the event). This was done to avoid clips with ambiguous meaning and ensure that there was general agreement on what was portrayed in each clip.

The data were analyzed in two steps. First the descriptions for all clips were compared and only messages that had more than 80% agreement on gist were selected. Then average positivity, negativity, and arousing content scores were computed for the remaining messages and, based on these scores, the messages were assigned to Valence × Arousing content conditions. Based on these ratings, the best three messages in each condition were then selected.

Defining detail

Details in this study are a stimulus property. They are defined as visual content of the message. Central details are characters or items that “make up” the gist; that is, they must be encoded to “derive” gist, to get the “basic interpretation” of the scene. Peripheral details are visual elements of the stimuli that do not have to be encoded to “derive” the gist. In other words, there would be no change in the basic interpretation of the scene if they were absent. Operational definitions of central and peripheral detail are provided in the following descriptions of recognition and cued recall test procedures.

Dependent Variables

Detail recognition

Initially, 48 still video frames were selected from each message for use as targets in the visual recognition test. One frame was selected from the first 30 seconds of the message and the other from the following 30 seconds. The frames used were at least 20 seconds apart. Each frame was then masked (as described below) to be used first in the central detail recognition and then in the peripheral detail recognition tests. Thus, a total of 24 recognition items for central detail and 24 items for peripheral detail were included in the forced-choice, speeded, yes/no recognition test. In addition, 48 (24 central and 24 peripheral) foils were selected from the same movies featuring the same main characters, but either in different appearances or involved with different actions in different contexts to form corresponding foil items for each type of detail.

There was one foil item for each target. Each recognition item was presented for 100 milliseconds, after which subjects were asked to respond “yes” if they remembered the objects or “no” if they thought the scene was new. Participants were encouraged to respond as quickly as possible. Central and peripheral items were blocked. The central detail recognition test was administered before the peripheral detail recognition test to reduce the possibility of a ceiling effect for central detail items. The order of the items within each block was randomized.

In order to ascertain that the selected frames contained sufficiently distinct peripheral and central detail, three coders were trained to code central and peripheral detail. During training, the coders were provided with definition of the central and peripheral detail and also with the frames preselected by the researcher from nonstimuli material. They were asked to look at each frame and identify up to four elements of the depicted event that, in their opinion, were central. They were then asked to identify up to four details that they considered peripheral. Results were then compared and disagreements solved by clarifying coding instructions and definitions. After this, each coder coded his or her portion of the frames. Frames were preselected by the researcher from stimulus material such that there were four frames from each message. The primary researcher specifically selected medium-long camera shots with a distinct foreground and background. The frames were distributed among coders so that each frame was coded by two coders. Only frames that achieved the highest agreement on central and peripheral detail were selected to be targets. Because coder agreement was used as a measure of stimulus reliability, assessment of intercoder reliability was not performed. In other words, the coders had to agree 100% before any detail could be selected for the test. Recognition items were masked. Two masks were designed using Adobe Photoshop software. One was a mask for central content the other was a mask for peripheral content. The central content mask consisted of one third of the entire picture area and had the same aspect ratio (ratio of width to height) as the original picture. This was achieved by measuring the height and the width of the original picture and then dividing it by the square root of three. This provided the height and the width of the area for the mask. The inverse of the mask area (the entire area of the original picture minus the central mask area) provided a mask for the periphery. These masks were then applied to the original picture. When the central mask was applied, only the peripheral aspects of the picture were visible. When the peripheral mask was applied, only the central part of the picture was visible. These masked pictures were used as the targets in the visual recognition test.

The same procedure as that used for creating masks was used to create foils. Pictures used as foils for the visual recognition test were selected to match the target thematically and in terms of visual complexity. There was one foil for each target.

Detail cued recall

The cued recall test consisted of 12 fill-in-the-blank items for each message. The questions required recalling objects or qualities (characteristics) of objects in a scene. There were four questions from the beginning, middle, and the end of each message. Two questions were about central objects and two were about objects from the periphery. Participants received a verbal description of the scene presented in the picture to cue them to the correct message and then were presented with the same statement, but with a missing word (blank). The task was to fill in the blank to make the statement a correct description of the object or scene as they remembered it. Participants were instructed to guess if they did not remember the answer. Questions for each message were presented in a block of four (two central and two peripheral). Within a block, questions for central objects were always presented first, followed by questions about periphery. The blocks were randomized by MediaLab software.

To ensure salience of both central and peripheral details for most of the participants, details selected for questions were based on results of a pretest during which five ICR lab researchers were provided with definitions of central and peripheral detail (as described earlier) and then were asked to list three central and three peripheral details from the first, middle, and last 20 seconds of each message. An effort was made to ensure that each message was evaluated by at least two people. These were then compared with the investigator’s selections. Only details that received agreement among the majority were selected for the cued recall test.

Procedure

Eighty participants were recruited from an introductory telecommunications course (35 female). Participants’ age ranged from 18 to 29 with an average equal to 21.62 (SD = 1.97). They were run in groups of 5 or 6 on individual laptop computers while wearing headphones. To ensure privacy, participants were separated from each other by dividers placed around the laptops. Each participant was instructed to watch each message as if he or she were doing so at home. After viewing, participants watched a 10-minute distracter video. They were then asked to complete the cued recall test, followed by the recognition test. Finally, they were thanked and dismissed.

Data Reduction and Analysis

Recognition

Four subjects’ data were lost due to subject withdrawal or equipment failure resulting in a final N = 76. The recognition assessment was based on a memory sensitivity measure d’, because this measure indicates how well people distinguish between old (targets) and new items (foils) independent of recognition memory decision criterion. This criterion is set by a person on a familiarity continuum (Shapiro, 1994) and reflects how familiar the item should be to a person, before it is recognized as old. If familiarity value for an item is larger than the criterion, a person recognizes the item as previously seen. However, if a personal criterion is set too low (is too liberal) an individual not only produces more hits (correct recognitions) but also risks production of a larger number of false recognitions (Macmillan & Creelman, 1991; Shapiro, 1994). Thus a recognition response “old” (yes, previously seen) may sometimes be more reflective of a liberal judgment criterion rather than memory strength. The signal detection measures such as d’ or criterion bias statistics allow researchers to focus on one aspect of recognition response at a time. D’ prime points to memory strength. Larger values of d’ indicate higher levels of memory strength.

Procedures described in Macmillan and Creelman (1991) were used to calculate d’.¹ Unless otherwise specified, an Arousing content (2) × Valence (2) × Detail (2) ANOVA model was run. Arousing content had two levels: low and high. The valence factor consisted of the positive and negative categories. The detail factor consisted of central item detail and peripheral item detail levels.

Cued recall

All incorrect responses and responses of “don’t know” were coded as “0”. Only correct responses were coded “1”. The response was considered a correct response if the answer was a match. The match was determined by the context of the event and by how closely the response reflected the nature or quality of the detail item that needed to be recalled. The final data set consisted of N = 81.

An Arousing content (2) × Valence (2) × Detail (2) × Message repetition (3) × Time (3) × Question repetition (2) ANOVA was run unless otherwise specified. Arousing content refers to two levels: low and high. The valence factor refers to the positive and negative categories. The detail factor consisted of central item detail and peripheral item detail levels. Message repetition refers to three messages in each arousing by content cell. Time factor refers to three frames chosen from the beginning, middle, and end of the clips, and question repetition refers to two questions for each type of detail.

Results

Hypothesis 1 predicted that recognition for negative content central and peripheral detail would be such that memory would not decrease for arousing content compared to that of calm. There was a significant arousing content by valence by detail interaction F(1, 75) = 4.46, p = .01, partial η² = .19 shown on Figure 1(A). A follow up Arousing Content (2) × Detail (2) ANOVA on negative content data produced a significant arousing content-by-detail interaction F(1, 75) = 8.5, p = .005, partial η² = .10, which showed a significant increase in recognition memory strength from low (M = 1.68, SD =.44) to high arousing content level (M = 2.94, SD =.44) for central detail and no decrease in peripheral content memory. This was also supported by follow-up pairwise comparisons t(75) = −3.98, p < .001 for central detail and t(75) = .72, p = .47 for peripheral detail.

Figure 1.

Recognition sensitivity d’ as a function of arousing content, valence, and detail

Hypothesis 2 predicted that recognition for positive content detail would be such that memory for detail would decrease for high arousing content compared to that of low, especially for peripheral detail. This was tested by an Arousing Content (2) × Detail (2) ANOVA on positive content. There was an interaction of arousing content and detail F(1, 75) = 8.78, p = .004, partial η² = .10 such that peripheral details for positive clips were identified worse in arousing (M = 1.07, SD =.49) compared to calm (M = 2.40, SD =.50) condition, yet there were no differences in recognition for central detail as a function of arousing content. This was supported by follow-up pair-wise comparisons t(75) = .61, p = .54 for central detail and t(75) = −4.27, p < .001 for peripheral detail. See also Figure 1(B).

Hypothesis 3 predicted that cued recall for negative content detail would be such that memory for detail would decrease for high arousing content compared to that of low, especially for peripheral detail. There was a significant arousing content by valence by detail interaction F(1, 80) = 4.38, p = .04, partial η² = .09. This interaction was further explored by an Arousing Content (2) × Detail (2) × Message repetition (3) × Time (3) × Question repetition (2) ANOVA on negative content data, which produced a significant arousing content by detail interaction F(1,80) = 6.0, p = .01, partial η² = .08. As illustrated in Figure 2(A), the interaction is driven primarily by a significant drop in peripheral detail recollection at high arousing content level. Peripheral details for negative clips were remembered worse in arousing (M = .32, SD =.02) compared to calm (M = .50, SD =.03) condition also supported by t-test comparison t(80) = 3.61, p < .001. There were no differences in recollection for central detail as a function of arousal t(80) = .43, p = .45.

Figure 2.

Cued recall accuracy as a function of arousing content, valence, and detail

Hypothesis 4 predicted that cued recall for positive content detail would be such that memory for both central and peripheral detail would not decrease for high arousing content level compared to that of low. An Arousing Content (2) × Detail (2) × Message repetition (3) × Time (3) × Question repetition (2) ANOVA on positive content data produced a significant arousing content by detail interaction F(1,80) = 13.85, p < .001, partial η² = .15 shown on Figure 2(B). Central details for positive clips were remembered better in arousing (M = .60, SD =.04) compared to calm (M = .53, SD = .03) condition also supported by a pairwise comparison t(80) = 4.23, p < .001. There were no differences in recall for peripheral detail as a function of arousing content t(80) = .63, p = .53.

Discussion

Overall, data support the initial assumption that visual recognition aids negative arousing content detail memory more than positive. In contrast, cued recall procedure is more helpful for recalling positive detail. For central detail, the matching cues work together with the arousing content to boost memory specificity, yet for peripheral detail the matching cues help overcome the harmful effects of arousing content by reducing forgetting. In sum, visual recognition minimizes the gun focus effect for negative content and enhances it for positive. At the same time, a cued recall task minimizes gun focus effect for positive visual content and enhances it for negative.

Theoretically, the findings support the notion that a link exists between encoding and retrieval mechanisms and that, when these are matched, remembering improves. Most importantly, the study calls for special attention to the role of valence factor in information processing. While a number of cognitive processing theories integrate emotion as an important factor, the majority of them (e.g., limited capacity, excitation transfer, source monitoring, consolidation) focus primarily on an arousal factor as a central determinant of attentional and memory effects. Emotion is conceived of as a powerful booster or barrier. Yet current advances in the understanding of the relationship between emotion and cognition (Cacioppo & Gardner, 1999; Kensinger, 2009; Lang, 2006; Mather, 2007) suggest that the valence dimension of emotional experience compliments the boosting/blocking effects of arousal by determining how the evoked cognitive activity is distributed. These findings are instrumental in advancing understanding of emotion−cognition interaction, and therefore are central in understanding processes beyond social communication behavior. Results here suggest that the valence factor plays an important role in determining not only the amount of information encoded but also how it is encoded and retrieved. The data are consistent with findings that suggest that gun focus can be eliminated under certain conditions. For example, Kensiger et al. (2007) were able to eliminate it for negative content when they specifically instructed participants to pay attention to the events as a whole, for later description purposes. The findings also make sense in light of the paradigm within which the gun focus was most often observed. The tasks in most of these studies employed either a cued recall or else verbal recognition scenarios. The fact that this study demonstrated memory narrowing for negative arousing content during cued recall is consistent with previous observations. The most important finding here is that peripheral information decay can be stopped if the right tools are employed.

The findings of this study have practical implications in a variety of communication scenarios. In eye-witness testimony, for example, different memory tasks can be useful depending on a valence of the arousing event being recalled. If the event is highly negative, recognition tasks may prove to be more fruitful than cued recall tasks and result in more accurate responses. In doctor−patient communication, to take a second example, when asking about extremely negative, painful experiences, doctors may be able to obtain more accurate and detailed information if they design interviews in a true/false format, as opposed to relying only on patient free recall.

In the arena of mediated content, however, viewers rely more on recall than on recognition. Advertising for a political candidate, for example, could be altered to stress a candidate’s character or position on particular issues (central details). In this case, creating an ad with a negative emotional tone would be more beneficial, as negative tone promotes recall of central detail and suppresses recall of peripheral. On the other hand, if contextual details, such as source of information or political affiliation, were to be stressed, then positive emotional tone would be more effective.

The study reported here provides only initial evidence for the modulating effects of emotion and retrieval cues on memory for detail. It is critical to keep in mind that the findings are limited to only a small pool of mediated content. It is yet to be determined whether the effects would hold for other genres, such as news or promotional messages or other types of detail such as verbal content and text. This is certainly a fruitful avenue for future exploration.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interests with respect to the authorship and/or publication of this article.

Funding

The author received no financial support for the research and/or authorship of this article.

Notes

Bio

Narine S. Yegiyan, PhD Indiana University, Bloomington, is an assistant professor in the Department of Communication at University of California, Davis. Her primary research interests include understanding of how message structure and content interact to affect cognitive and emotional processing of media. She has published in such journals as Human Communication Research, Media Psychology, and Cognition & Emotion.

References

Bernstein

L. J.

Berg

Siegenthaler

A. L.

Grady

C. L.

(2002). The effect of encoding strategy on the neural correlates of memory for faces. Neuropsychologia, 40, 86-98.

Bolls

P. D.

Lang

Potter

R. F.

(2001). The effects of message valence and listener arousal on attention, memory, and facial muscular responses to radio advertisements. Communication Research, 28, 627-651.

Bradley

S. D.

Angelini

J. R.

Lee

(2007). Psychophysiological and memory effects of negative political ads: Aversive, arousing, and well remembered. Journal of Advertising, 36(4), 115-127.

Buchanan

T. W.

Adolphs

(2002). The role of the human amygdala in emotional modulation of long-term declarative memory. In Moore

Oaksford

(Eds.), Emotional cognition: From brain to behavior (pp. 9-34). Amsterdam, Netherlands: John Benjamins.

Bucy

E. P.

Newhagen

J. E.

(1999). The emotional appropriateness heuristic: Processing televised presidential reactions to the news. Journal of Communication, 49(4), 59-79.

Cacioppo

J. T.

Gardner

W. L.

(1999). Emotion. Annual Reviews: Psychology, 50, 191-214.

Christianson

S. A.

Loftus

(1991). Remembering emotional events: The fate of detailed information. Cognition & Emotion, 5, 81-108.

Clifford

B. R.

Hollin

C. R.

(1981). Effects of the type of incident and the number of perpetrators on eyewitness memory. Journal of Applied Psychology, 66, 364-370.

Craik

F. I. M.

Lockhart

R. S.

(1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671-684.

10.

Dobbins

I. G.

Wagner

A. D.

(2005). Domain-general and domain-sensitive prefrontal mechanisms for recollecting events and detecting novelty. Cerebral Cortex, 15, 1768-1778.

11.

Fredrickson

B. L.

Branigan

(2005). Positive emotions broaden the scope of attention and thought-action repertoires, Cognition & Emotion, 19, 313-332.

12.

Garoff

R. J.

Slotnick

S. D.

Schacter

D. L.

(2005). The neural origins of specific and general memory: The role of the fusiform cortex. Neuropsychologia, 43, 847-859.

13.

Gasper

Clore

G. L.

(2002). Attending to the big picture: Mood and global vs. local processing of visual information. Psychological Science, 13, 34-40.

14.

Goss

Neuliep

J. W.

O’Hair

(1985). Reduced conversational memory as a function of negative arousal: A preliminary analysis. Communication Research Reports, 2, 202-205.

15.

Grabe

M. E.

Kamhawi

(2006). Hard wired for negative news? gender differences in processing broadcast news. Communication Research, 33, 346-369.

16.

Heuer

Reisberg

(1990). Vivid memories of emotional events: The accuracy of remembered minutiae. Memory & Cognition, 18, 496-506.

17.

Kahn

Davachi

Wagner

A. D.

(2004). Functional-neuroanatomic correlates of recollection: Implications for models of recognition memory. Journal of Neuroscience, 24, 4172-4180.

18.

Kebeck

Lohaus

(1986). Effect of emotional arousal on free recall of complex material. Perceptual and Motor Skills, 63, 461-462.

19.

Kensinger

E. A.

(2009). Remembering the details: Effects of emotion. Emotion Review 1(2), 99-133.

20.

Kensinger

E. A.

Garoff-Eaton

R. J.

Schacter

D. L.

(2007). Effects of emotion on memory specificity: Memory trade-offs elicited by negative visually arousing stimuli. Journal of Memory and Language, 56, 575-591.

21.

Kirchhoff

B. A.

Wagner

A. D.

Maril

Stern

C. E.

(2000). Prefrontal-temporal circuitry for episodic encoding and subsequent memory. Journal of Neuroscience, 20, 6173-6180.

22.

Kuskowski

M. A.

Pardo

J. V.

(1999). The role of the fusiform gyrus in successful encoding of face stimuli. Neuroimage, 9, 599-610.

23.

Lang

(2006). Using the limited capacity model of motivated mediated message processing to design effective cancer communication messages. Journal of Communication, 56, S57-S80.

24.

Lang

Dhillon

Dong . (1995). The effects of emotional arousal and valence on television viewers’ cognitive capacity and memory. Journal of Broadcasting and Electronic Media, 39, 313-328.

25.

Lang

Newhagen

Reeves

(1996). Negative video as structure: Emotion, attention, capacity, and memory. Journal of Broadcasting & Electronic Media, 40, 460-477.

26.

Libkuman

T. M.

Stabler

C. L.

Otani

(2004). Arousal, valence, and memory for detail. Memory, 12, 237-247.

27.

Loftus

E. F.

Burns

T. E.

(1982). Mental shock can produce retrograde amnesia. Memory & Cognition, 10, 318-323.

28.

Macmillan

Creelman

(1991). Detection theory: A user’s guide. Cambridge, UK: Cambridge University Press.

29.

Mather

(2007). Emotional arousal and memory binding: An object based framework. Perspectives on Psychological Science, 2, 33-52.

30.

Mickley

K. R.

Kensinger

E. A.

(2008). Emotional valence influences the neural correlates associated with remembering and knowing. Cognitive, Affective, and Behavioral Neuroscience, 8, 143-152.

31.

Nyberg

Habib

McIntosh

A. R.

Tulving

(2000). Reactivation of encoding-related brain activity during memory retrieval. Proceedings of the National Academy of Sciences USA, 97, 11120-11124.

32.

Poldrack

R. A.

Selco

S. L.

Field

J. E.

Cohen

N. J.

(1999). The relationship between skill learning and repetition priming: Experimental and computational analyses. Journal of Experimental Psychology:Learning, Memory, and Cognition, 25, 208-235.

33.

Reisberg

Heuer

(2004). Memory for emotional events. In Reisberg

Hertel

(Eds.), Memory and emotion (pp. 3-41). New York, NY: Oxford University Press.

34.

Rowe

Hirsh

J. B.

Anderson

A. K.

(2007). Positive affect increases the breadth of attentional selection. Proceedings of the National Academy of Sciences USA, 104, 383-388.

35.

Shapiro

M. A.

(1994). Signal detection measures of recognition memory. In Lang

(Ed.), Measuring psychological responses to media messages (pp. 133-148). Hillsdale, NJ: Lawrence Erlbaum.

36.

Shiffrin

R. M.

(1999). 30 years of memory. In Izawa

(Ed.), On human memory: Evolution, progress, and reflections on the 30th anniversary of the Atkinson-Shiffrin model (pp. 17-33). Mahwah, NJ: Lawrence Erlbaum.

37.

Shoemaker

P. J.

(1996). Hardwired for news: Using biological and cultural evolution to explain the surveillance function. Journal of Communication, 46(3), 32-47.

38.

Talarico

J. M.

Berntsen

Rubin

D. C.

(2009). Positive emotions enhance recall of peripheral details. Cognition and Emotion, 23(2), 380-398.

39.

Vaidya

C. J.

Zhao

Desmond

J. E.

Gabrieli

J. D.

(2002). Evidence for cortical encoding specificity in episodic memory: Memory-induced re-activation of picture processing areas. Neuropsychologia, 40, 2136-2143.

40.

Wheeler

M. E.

Petersen

S. E.

Buckner

R. L.

(2000). Memory’s echo: Vivid remembering reactivates sensory-specific cortex. Proceedings of the National Academy of Sciences USA, 97, 11125-11129.

41.

Yegiyan

N. S.

Lang

(2010). Processing central and peripheral detail: How content arousal and emotional tone influence encoding. Media Psychology, 13, 77-99.

42.

Yegiyan

N. S.

Yonelinas

A. P.

(2011). Encoding details: Positive emotion leads to memory broadening. Cognition & Emotion, 25(7), 1255-62.