<%BANNER%>

Choking under Pressure and Working Memory

Permanent Link: http://ufdc.ufl.edu/UFE0041883/00001

Material Information

Title: Choking under Pressure and Working Memory Influence of Cognitive Appraisal
Physical Description: 1 online resource (43 p.)
Language: english
Creator: Wang, Ye
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Educational Psychology thesis, M.A.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Eysenck (1992) proposed that anxiety could influence cognition by occupying cognitive processing resources, jeopardizing working memory capacity by pressure via increasing state anxiety. Recent findings (Beilock & Carr, 2001, 2005; Gimming, Huguet, Caverni, & Cury, 2006) have demonstrated that only individuals with high working memory (HWM) capacity ?choke under pressure? on cognitive tasks with high working memory demands; while individuals with low working memory (LWM) capacity increased their performance under such conditions. Gimming and his colleagues suggested that LWM and HWM individuals interpreted stress differently, but research has not explored how these groups may be different. Our present study examined how LWM and HWM individuals interpreted stress differently by applying cognitive appraisal theory (Lazarus and Folkman, 1984). Our results suggested that HWM group experienced highest threat appraisal and thus choked under pressure.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ye Wang.
Thesis: Thesis (M.A.E.)--University of Florida, 2010.
Local: Adviser: Therriault, David.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041883:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041883/00001

Material Information

Title: Choking under Pressure and Working Memory Influence of Cognitive Appraisal
Physical Description: 1 online resource (43 p.)
Language: english
Creator: Wang, Ye
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: Human Development and Organizational Studies in Education -- Dissertations, Academic -- UF
Genre: Educational Psychology thesis, M.A.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Eysenck (1992) proposed that anxiety could influence cognition by occupying cognitive processing resources, jeopardizing working memory capacity by pressure via increasing state anxiety. Recent findings (Beilock & Carr, 2001, 2005; Gimming, Huguet, Caverni, & Cury, 2006) have demonstrated that only individuals with high working memory (HWM) capacity ?choke under pressure? on cognitive tasks with high working memory demands; while individuals with low working memory (LWM) capacity increased their performance under such conditions. Gimming and his colleagues suggested that LWM and HWM individuals interpreted stress differently, but research has not explored how these groups may be different. Our present study examined how LWM and HWM individuals interpreted stress differently by applying cognitive appraisal theory (Lazarus and Folkman, 1984). Our results suggested that HWM group experienced highest threat appraisal and thus choked under pressure.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Ye Wang.
Thesis: Thesis (M.A.E.)--University of Florida, 2010.
Local: Adviser: Therriault, David.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041883:00001


This item has the following downloads:


Full Text





CHOKING UNDER PRESSURE AND WORKING MEMORY:
INFLUENCE OF COGNITIVE APPRAISAL





















By

YE WANG


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS IN EDUCATION

UNIVERSITY OF FLORIDA

2010




































2010 Ye Wang





























To Jie Zou









ACKNOWLEDGMENTS

I wish to thank several individuals for their help and support concerning my thesis study.

First, I thank Dr. David Therriault, my advisor and the chair of my thesis committee, for

providing me with generous support throughout the process. His constructive and prompt

feedback on my study design, data analyses, and writing was greatly appreciated. I also thank the

other member of my thesis committee, Dr. David Miller, for his service and helpful comments

about my study. Last but not the least, I acknowledge Jie Zou, my husband, for supporting and

encouraging me as well as helping me with the writing style. I thank my parents for their love

and encouragement, and unconditional support for my interests and dreams. All these individuals

have helped me achieve a goal that seemed like a very daunting challenge.










TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

ABSTRAC T .......................................................................... 7

CHAPTER

1 O V E R V IE W .............. ................................. ............ ................................... 8

2 M E T H O D S ....................... ..................... .......................................................................... 17

S a m p le ............................................................................1 7
M ate ria ls ...........................................................................1 7
B ell C u rv e S u rv ey ..................................................................................................... 17
W working M em ory T est ..............................................................17
S elf-H an dicap p in g S cale ........................................................................................... 19
D em graphic Q questionnaire ...................................................................................... 19
State Appraisal Measure .................................... ........................... ..........19
Self-Reported Appraisal and Coping Expectancies ................. ................. ....20
State A anxiety Scale........................................................ 21
P ro c e d u re .............. .... ............................................................... 2 1

3 R E SU L T S .............. ... ................................................................24

R ead in g S p an T a sk ........................................................................................................... 2 4
State A appraisal Score.........................................................26
Self-R reported A appraisal Scale........................................................27
State Anxiety Scale.............................................. 28
S tre ss S c o re ........................ ............................................................................................... 2 8
E effort Score...... ..................................................28
Self-H handicapping Scale ................................................................29
D em graphic Q u estionnaire ............................................................................................. 29

4 D ISC U SSIO N ......................................................................................................................... 30

APPENDIX

A DEMOGRAPHIC QUESTIONNAIRE ................................. .......................... ....36

B STA TEM EN T Q U ESTION N A IRE ................................................................................. 37

C S C R IP T ........................ .................... ................................................................................ 3 8

LIST OF REFERENCES .............................39..............................................









B IO G R A PH IC A L SK E T C H ............................................................................... .....................43









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts in Education

CHOKING UNDER PRESSURE AND WORKING MEMORY:
INFLUENCE OF COGNITIVE APPRAISAL

By

Ye Wang

August 2010

Chair: David Therriault
Major: Educational Psychology

Eysenck (1992) proposed that anxiety could influence cognition by occupying cognitive

processing resources, jeopardizing working memory capacity by pressure via increasing state

anxiety. Recent findings (Beilock & Carr, 2001, 2005; Gimming, Huguet, Caverni, & Cury,

2006) have demonstrated that only individuals with high working memory (HWM) capacity

"choke under pressure" on cognitive tasks with high working memory demands; while

individuals with low working memory (LWM) capacity increased their performance under such

conditions. Gimming and his colleagues suggested that LWM and HWM individuals interpreted

stress differently, but research has not explored how these groups may be different. Our present

study examined how LWM and HWM individuals interpreted stress differently by applying

cognitive appraisal theory (Lazarus and Folkman, 1984). Our results suggested that HWM group

experienced highest threat appraisal and thus choked under pressure.









CHAPTER 1
OVERVIEW

Working memory plays an essential role in everyday life; it helps us to keep track of where

we are and what we are doing moment to moment. It aids us in holding information long enough

to make a decision. Storing while simultaneously processing information is fundamental to many

cognitive tasks such as reasoning, reading comprehension and learning (Baddeley, 2002).

Working memory capacity (WMC) is also limited. Individuals with high working memory have

greater resources at their disposal to control processing and maintain information. Controlled

processing is required for gating, blocking, or suppressing distracting information in the face of

distraction and interference in order to maintain temporary goals (Engle, Kane, & Tuholski,

1999).

There are many potential influences on working memory. For example, working memory

is susceptible to interference and intrusions. One source of interference and intrusion comes from

anxiety, or stress. Eysenck (1992) proposed that anxiety could influence cognition by occupying

cognitive processing resources and WMC was jeopardized by pressure via increasing state

anxiety (Beilock & Carr, 2001, 2005; Beilock, Tugade, & Engle, 2004). People exposed to stress

showed significant working memory impairments (Schoofs, 2008). Ashcraft (2002) found math

anxiety disrupted cognitive processing by compromising ongoing activity in working memory.

Ashcraft and Kirk (2001) found individuals with high math anxiety demonstrated smaller

working memory spans. They considered disruption of central executive processes as a possible

mechanism and suggested individual emotional difference variables like math anxiety deserve

greater attention, especially on assessments of working memory capacity and functioning.

One research approach to induce anxiety and stress in laboratory settings is by using

stereotype threat. According to stereotype threat theory (Steele & Aronson, 1995), performing in









a negatively stereotyped group produces feelings of anxiety, uncertainty, and discomfort.

Stereotype threat often impairs performance (Aronson, Lustina, Good, Keough, Steele, &

Brown, 1999; Croizet & Claire, 1998; Steele & Aronson, 1995). Considerable research has

indicated that cognitive and affective consequences of activating a negative self-relevant

stereotype of race (Cheryan & Bodenhausen, 2000), social economic statues (Steele, 1997), or

gender (Spencer, Steele, & Quinn, 1999; Steele, 1997), including decreased performance

expectations (Cadinu, Maass, Frigerio, Impagliazzo, & Latinotti, 2003; Stangor, Carr, & Kiang,

1998), increased anxiety (Bosson, Kaymovitz, & Pinel, 2004; Osborne, 2001), and reduced

working memory capacity (Beilock, Rydel, & McConnell, 2007; Schmader & Johns, 2003).

Stress can lead to "choking." "Choking under pressure" is defined as performing more

poorly than expected given one's skill and occurs in situations where the pressure to perform is

high (Beilock & Carr, 2001). Having examined the relationship between choking under pressure

and individual differences in working memory capacity, some authors concluded that the

individuals who were most likely to fail under performance pressure were those who, in the

absence of pressure, had the highest potential for success (i.e., high working memory

individuals). Beilock and Carr (2005) found that only high working memory (HWM) individuals

were harmed by performance pressure in math problems. Gimmig, Huguet, Cavemi and Cury

(2006) also found that individuals with high working memory were more likely than their low-

working-memory (LWM) counterparts to choke under pressure on working memory and fluid

intelligence tests. In Lai's (2008) results, individuals low in working memory capacity (when

given negative feedback about their performance) actually improved on working memory tasks.

In contrast, HWM individuals who were given the same feedback demonstrated choking.









Several potential reasons for choking under pressure have been proposed. Anxiety is the

one that has been most explored (Beilock & Carr, 2001; Beilock & Carr, 2005; Gimming et al.,

2006). Beilock and Carr (2001) explained that heightened anxiety induced self-focused attention,

thus producing intrusive thoughts. Beilock and Carr (2005) explained that performance pressure

increased reported state anxiety. Intrusive thoughts, such as worries, reduced the storage and

processing capacity of working memory, anxiety-providing situations reduced attentional

resources, resulting in performance decrements. They explained that HWM individuals' usual

working memory advantage might make them susceptible to failure when pressure was added, if

pressure-induced consumption of working memory denied them the capacity they normally

relied on to produce their superior performance. In line with Beilock and Carr (2005), Gimming

and his colleagues (2006) found anxiety mediated choking among HWM individuals, and

choking under pressure was due to increased state anxiety for HWM individuals. Interestingly,

they found under same pressure situation as HWM individuals' experience, LWM individuals

reported decreased anxiety and increased performance. They suggested that LWM and HWM

individuals might interpret high-stress situations differently. Anxiety increased in HWM

individuals under pressure, but decreased in LWM individuals. They explained that only HWM

individuals choked because of their anxious perception of high-stakes situations at the onset.

In sum, reactions to pressure appear to depend on WMC. It appears that choking only

occurs in HWM individuals. Previous studies are less clear about the specific mechanism driving

these results. According to Eysenck (1992), anxiety occupies cognitive processing resources. The

question remains, why is this effect only occur for HWM individuals. In fact, LWM individuals

demonstrated decreased anxiety (Gimming, et al, 2006) and even increased performance (Lai,

2008) under high pressure situation. Gimming and his colleagues (2006) suggested that the









pressure also affected LWM individuals, but their performance increased in pressure situation. If

anxiety compromises working memory capacity, why doesn't the same pattern of choking occur

for LWM individuals? The purpose of our study was to further explore the relationship of

anxiety and working memory. Specifically, we examined the link between working memory and

stress by employing cognitive appraisal theory.

In cognitive appraisal theory, stress can have differential effects. The effect of stress can be

either positive or negative. Negative effect of stress causes anxiety and decreases performance,

referred as threat appraisal. Positive effect of stress increases performance on a task causing what

is referred as challenge appraisal (Lazarus & Folkman, 1984). For example, Vick, Seery,

Blascovich and Weisbuch (2008) found an effect of gender stereotype activation related to

challenge or threat cognitive appraisal states. Girls, who were told that experiment was gender

biased, would feel gender stereotype threat; therefore they exhibited the threat state and

underperformed. If girls were told that experiment was gender fair, they would not feel

stereotype threat; instead, they exhibited challenge appraisal and perform better. In contrast, if

the experimenter told boys that the experiment was gender fair, they showed threat and

underperformed. If boys were told that experiment was gender biased (i.e., females would

underperform), they would show the challenge state and increase performance. Thus, studies

regarding cognitive appraisal theory suggested that interpretation of stress could either increase

or decrease performance, depending on the appraisal of the stress.

In cognitive appraisal theory, Lazarus and Folkman (1984) proposed two major

components of the processes: primary appraisal and secondary appraisal. Primary appraisal

reflects perceptions of the nature and degree of risk in the situation. Secondary appraisal reflects

perceptions of resources or abilities to cope with the situation. According to appraisal theories,









before performing a task, the task is appraised on two criteria: "how difficult the task will be"

and "how well I could cope with the task demands." Appraisals can then be classified as either

"threatening" or "challenging." Threat appraisals are those in which the perception of danger

exceeds the perception of abilities or resources to cope with the stressor, and result in

underperformance. Challenge appraisals, in contrast, are those in which the perception of danger

does not exceed the abilities to cope, and possibly increase performance. In a serial subtraction

tasks (i.e., the participant was asked to repeatedly subtract a small number, like 7 or 13, from a

running total that starts as a four-digit number, such as 5,342), threatening appraisals leaded to

performances with more mistakes and about a 25% slower rate than the challenging appraisal

conditions (Tomaka, Blascovich, Kelsey, & Leitten, 1993).

Threat appraisals are more strongly associated with negative emotional reactions, high

rates of cognitive interference, and poor performance levels than challenge appraisals are.

Threatened individuals may be less task-focused and have intrusive cognitions, whereas

challenged individuals may be less distracted by negative emotions and more task-focused.

Folkman and his colleagues (1986) also found that participants who appraised self-reported

stressful events as changeable (high secondary appraisal of coping efficacy) used task-focused

coping strategies (e.g., planed problem solving), and were more motivated and more satisfied

with encounter outcomes than participants who appraised the events as unchangeable. These

participants (low secondary appraisal of coping efficacy) used more emotion-focused coping

strategies (e.g., distancing and escape-avoidance). These emotional coping strategies left

participants less motivated to finish the task, and more likely to try to find self-handicapping

excuses. Skinner and Brewer (2002) demonstrated that challenge appraisals were associated with

more confident coping expectancies, lower perceptions of threat, higher positive emotion, and









more beneficial perceptions of the effects of appraisal and emotion on performance. Generally,

higher performance was associated with challenge appraisals. Gaab, Rohleder, Nater, and Ehlert

(2005) linked appraisal with psychobiological stress change and found that appraisal helped to

predict performance. Vick and his colleagues (2008), as we reviewed above, found an effect of

gender stereotype activation due to challenge/threat states. Females in the gender-biased

condition exhibited more threatened state, whereas women in the gender-fair condition exhibited

more challenge state than they did in the gender-biased condition. Thus, challenge and threat

states can be activated by anxiety of negative information of oneself and change performance.

Threatened individuals may be less task-focused, and more self-emotion focused, whereas

challenged individuals may be less distracted by negative emotions, and more task-focused. As

reviewed above, in working memory research, Beilock and Carr (2001) found self-consciousness

training, which trained participants not to focus on thoughts about self but focus on task,

eliminated choking. Intrusive thoughts, which are less task-focused, can cause choking (Beilock

and Carr, 2005). Heightened anxiety induces self-focused attention and causes choking. Given

our review of the literature, it was reasonable to hypothesize that choking is related to cognitive

appraisal. Cognitive appraisal might help to explain why choking works differently in high or

low working memory individuals. It is possible that HWM people are more likely to demonstrate

threat appraisal and choke under pressure; whereas LWM people are more likely to demonstrate

challenge appraisal and avoid choking.

Furthermore, because individuals with threat appraisal lose the confidence to cope with the

problems they will be more likely to try to employ excuses to avoid the stress, resulting in high

self-handicapping scores. In contrast, for challenge appraisal individuals, they are likely to be

more motivated and demonstrate low self-handicapping scores under stressful situation.









Gimming et al. (2006) found differences in the self-handicapping measurements for HWMs and

LWMs. Choking was not only related to state anxiety, but also self-handicapping. Self-

handicapping was used to describe the tendency of trying not to focus on task in order to use this

as an excuse to avoid pressure. Gimming et al. (2006) reported that LWM people had less

anxiety score on state anxiety scale, and low self-handicapping score on Rhodewalt's scale

(1990), explaining why LWMs did not choke under pressure. In contrast, HWM individuals

displayed a tendency to excuse their poor performance under pressure (they scored higher in self-

handicapping than LWM individuals).

Thompson (2007) found that poor performance feedback could produce self-handicapping

behavior in people with high ability, which is analogous to the HWM individuals in our study. In

our study, we predicted that for LWMs, challenge appraisal would be linked to lower self-

handicapping scores; and for HWMs, threat appraisal would be linked to higher scores of self-

handicapping. Therefore, we incorporated in our study a self-handicapping scale (Rhodewalt's

scale, 1990) at the end of second task and Gimming et al. used the same scale in their study.

Previous studies focused on reasons why HWM individuals' choke. Relatively few studies

explored why LWM individuals perform better and seem to be under less anxiety when placed in

high-pressure situations. In Lai's (2008) study, she argued that LWM individuals did not use full

cognitive resources under low stressful task, saving cognitive resources for more stressful tasks.

In our study, we would also explore this hypothesis by giving a self-report effort scale at the end

of each task. On the other hand, this self-report effort scale could also give us important

information about whether HWM individuals' self-handicapping behavior would cause a

withdrawal of effort under high pressure.









The current study extends previous research by examining a potential cognitive appraisal

consequence in HWMs and LWMs. The current study was a 2 (WMC: HWMs vs. LWMs) x 2

(stress: treatment vs. control) design. Our study adopted Lai's (2008) approach of using feedback

to induce stress. Participants were categorized into high, middle or low working memory span

(while dropping the middle span group in the paired comparison analyses because it was not of

interest to our hypotheses). Participants were administered a working memory test, then given

feedback on how well they did by assigning them randomly into one of two groups/conditions:

40th percentile, or no-feedback condition. As noted by Beilock and Carr (2005), pressure that

caused people to choke included the consequences of suboptimal performance, especially on

examinations, including poor evaluations by mentors, teachers and peers. The 40th percentile

condition was expected to create stress for the second task (i.e., the participants were informed

that they received a failing score of 60%). Then participants then completed a second working

memory task. Participants in the no-feedback condition continued on to the second test without

any explicit feedback.

We predicted that people with HWM would be more likely to interpret feedback as threat

appraisal and choke under pressure while LWM people would not. Our specific hypotheses

regarding individuals in the 40th percentile feedback treatment condition were as follows: (1)

challenging participants' views of their own competency would result in choking, especially in

HWM individuals, not in LWM ones; (2) HWM people would demonstrate more state anxiety

than LWM people; This hypothesis could potentially provide confirmatory evidence for Beilock

and Carr's (2005) and Gimming et al.'s (2006) speculations that choking is mediated by anxiety;

(3) HWM individuals would show more threat appraisal in the second task than in the first task;

(4) HWMs would feel more stressful, put less effort and express high self-handicapping in the









second task than in the first task; (5) LWMs would show more challenge appraisal in the second

task than in the first task; (6) LWMs would feel less stressful, put more effort and express lower

self-handicapping in the second task than in the first task. In no feedback condition, we

hypothesized: (7) the control groups would have no significant changes in state anxiety,

cognitive appraisal, RSPAN performance, stress, effort in first and second task.









CHAPTER 2
METHODS

Sample

There were 157 participants in this study. They were undergraduate educational

psychology students from the University of Florida who participated for credits toward a course

requirement.

Materials

Bell Curve Survey

Upon signing the informed consent, participants were administered a Bell Curve Survey

asking them to indicate where they believed their academic competency (percentile) fell on the

curve, comparing to their peers. Because in our experiment, participants were informed that this

experiment correlated well with academic competency, the survey served to set the stage for the

manipulation of giving negative feedback treatment by highlighting social comparisons between

individuals.

Working Memory Test

We employed a dual-processing test called the reading-span task (RSPAN) to measure

working memory capacity (Conway, Kane, Bunting, Hambrick, Wilhelm, & Engle, 2005). The

idea is that high working memory capacity individuals do better job in dual tasks because they

have more resources to maintain and keep track of multiple sources of information. Daneman

and Carpenter (1980) originally developed the RSPAN task to assess working memory during

reading. In our study, participants evaluated sentences while memorizing letters for later recall so

it required dual processing. Participants were asked to read a sentence presented on the computer

screen, and decided whether or not that sentence made sense in a real world scenario by clicking

either "yes" or "no" button on the screen. A letter that required participants to recall later was









presented for 2 seconds after each sentence decision and a blank screen lasting 1 second

separated the presentation of each sentence and letter. At the end of a series of sentence/letter

combination trials, participants were asked to recall, as best as they could, the letters in the order

presented. Each set was separated by a screen said "next set", which was displayed for 3

seconds. The computer recorded the number of letters recalled in the correct order, and

participants' correct or incorrect sentences decision. One set ranged from three to six sentences,

so at the end of a set participants were asked to recall the letters he/she could recall in the set,

and then came to the next set.

In our study, the RSPAN task included 15 sets of sentence/letter trials that contained three

to six sentence/letter combinations. The sets were presented in random order so participants

could not anticipate how many sentences and letters they were required to evaluate and recall at

the beginning of each set. This was a dual task because participants made judgments whether the

sentence made sense, and at the same time remembered the letters presented after each sentence.

Working memory capacity was scored by the total number of letters that participants recalled

correctly from each set while maintaining at least an 85% accuracy rate on judging the sensibility

of sentences.

The second RSPAN task mirrored the first. Both tasks had equal number of sets and equal

number of nonsense and sense sentences. The letters to be recalled were also taken from the

same sampling of letters as the first RSPAN task. The sentences from the second RSPAN were

of equal length to the first task and same number of words per sentence. And the sentences are

also of similar difficulty levels using the Flesch-Kincaid Grade Level readability scale.

We chose RSPAN task because it was stressful and difficult, as research reports had

indicated (Conway et al., 2005; Lai, 2008). Stress is most likely to occur when the task is









sufficiently difficult to frustrate participants' abilities (Steele, Spencer, & Aronson, 2002).

Therefore, this task is likely to induce threat and cognitive appraisal of stress.

Self-Handicapping Scale

Participants were administered a self-handicapping scale after completing the second task.

They rated statements about how handicapped they had felt during the task performance on a 5-

point scale ranging from 0 (totally disagree) to 5 (totally agree) (e.g., "I slept well last night";

adapted from Rhodewalt, 1990). This served to index self-handicapping (a=0.69).

Demographic Questionnaire

Participants were administered a brief Demographic Questionnaire (Appendix A) at the

end of the study. This was included to insure that there was no accidental priming of a gender or

race stereotype threat that could affect performance. The questionnaire asked questions regarding

age, gender, and ethnicity. The questionnaire was included as an additional control measure.

State Appraisal Measure

Skinner and Brewer's (2002) reports provided an example of this scale used. Four threat

appraisal items addressed concerns about personal evaluation ("I'm concerned that others will be

disappointed in my performance") and performance ("I worry that I may not be able to achieve

the grade I am aiming for"; "I'm concerned about my ability to perform under pressure") and the

tendency to focus on negative outcomes ("I am thinking about the consequences of performing

poorly"); The other four challenge appraisal items addressed participants' enthusiasm toward

testing skills and abilities in demanding situations (" I am looking forward to testing my

knowledge, skills, and abilities") and the anticipation of positive benefits (" I am focused on the

positive benefits I will obtain from this situation"; I am looking forward to the rewards of

success"; "I am thinking about the consequences of performing well"). Intensity and frequency

of state appraisals were measured on separate 6-point response scales (1 = strongly disagree, 6 =









strongly agree and 1 = this thought hardly ever occurred, 6 = this thought occurred almost

constantly).

Self-Reported Appraisal and Coping Expectancies

This scale was similar to state appraisal scale because they both measure cognitive

appraisals. However, the self-reported appraisal scale asked about "the current state" (i.e., our

manipulation) because it asked participants about the upcoming task. In contrast, the state

appraisal scale is a more global assessment of general state rather than participants' feelings

about the specific task. Tomaka et al. (1993) assessed primary and second appraisal by asking

subjects, "How stressful do you expect the upcoming task to be?" Secondary appraisal was

assessed by asking: "How able are you to cope with this task?" before the task; and assessed

subjective experience of stress by asking, "How stressful was the task you just completed?" after

the task. Skinner and Brewer (2002) also measured coping expectancies by self-reported

anticipated coping ability and expected level of performance. Coping expectancies were assessed

by two 10-point scales tapping anticipated coping ability (1 = very confident can cope

effectively, 10 = very concerned whether can cope effectively) and the expected level of

performance (1 = poor, 10 = high quality). The second coping item was reverse coded. Kelsey et

al. (2000) also asked five questions immediately before performing task, which were (a) "How

threatening do you think the upcoming task will be?" (b) "How demanding do you think the

upcoming task will be?" (c) "How stressful do you think the upcoming task will be?" (d) "How

able are you to cope with this task?" and (e) "How well do you think you will perform this task?"

Obviously (a) was related to threat appraisal; (b) was related to challenge appraisal; (c) was the

primary appraisal; (d) and (e) were like what Skinner and Brewer used as secondary appraisal

coping expectancy.









We adopted the five questions from Kelsey's (2002) report and we used 10-point scales for

each question. These appraisal items were presented to subjects through computer after they had

read the instructions of the task. Finally we asked participants "How stressful was the task you

just completed?" after they had completed each of the tasks.

State Anxiety Scale

Following each task, participants completed an anxiety scale that has several items related

to perceptions of the test and testing situation. Anxiety during the tests was measured using the

questions adapted from the Spielberger State Anxiety Scale (1970) (see Schmader & Johns,

2003). Participants rated on a 7-point scale (ranging from 1 not at all to 7 very much) how much

they felt anxious, comfortable, jittery, worried, at ease, nervous, relaxed, and calm while taking

the test. The items on the scale were summed (after reverse scoring comfortable, at ease, relaxed,

and calm) to form an index of state anxiety, where higher numbers indicated more anxiety (a=

0.90).

Procedure

After seating a participant by a computer in a quiet laboratory room, a research assistant

first asked the participant to read and sign an informed consent form. The research assistant then

presented the Bell Curve Survey in which participants indicated their academic competency in

percentile compared to their peers on the bell curve. Participants were told that the experiment

consisted two tests and took about 1 hour long. The research assistant highlighted the importance

and credibility of the test by saying that the test was created by an authority and was related to

the academic performance. The research assistant also described working memory capacity so

that the participants were made fully aware of the task. The detailed script was included in

Appendix C. The first RSPAN test instruction was then presented on the computer. Immediately

after viewing and fully understanding the instruction, participants were measured on the state









appraisal measure and self-report appraisal and coping expectancies. Then the first RSPAN test

was administered.

Following the first test, participants were required to answer, "How stressful was the task

you just completed?" "How much effort did you put in?" (both questions were 10-point scale),

and the state anxiety scale. After completing the questions, the research assistant told them their

"performance" and indicated the percentage in the Bell Curve. Participants were randomly

assigned into one of two conditions/groups, which was no feedback, or negative (stress)

feedback. We chose to use percentile ranks as it highlighted the participant's own performance

compared to others. Even though our participants should have had also familiar with this type of

reporting which they experienced in standardized tests, like the SAT, the research assistant still

explained the percentile rankings and informed participants how their scores were compared to

others. Participants in the 40th percentile were told that based on their rank they were expected to

have difficulty on the subsequent test. The 40th percentile was used because it was considered

below average so as to create stress, but it was still a realistic achievement group. In the no-

feedback condition, participants completed the questionnaires and questions but continued onto

the second test without any score or expectation feedback. Using the no-feedback condition as

the control group, we would learn whether the two different RSPAN tasks were of equal

difficulty level.

The second RSPAN test was then administered. After the computer displayed the

instructions, participants were assessed on the same self-report appraisal and state appraisal

scales. Then the second RSPAN test started. After finishing the second RSPAN test, participants

were assessed on the same questions and state anxiety scales as they did in the first RSPAN test.

Upon completing above tests and scales for all participants, the experimenter let participants fill









out the self-handicapping scale, and a questionnaire asking how they felt after they received the

feedback, and how they perceived their performance, expectancy, effort and motivation changed

in the second task (Appendix B). Participants were then asked to fill out the demographic

questionnaire.

Finally, participants in the treatment condition were debriefed by informing participants

that their percentile rankings feedbacks for the first test were fictitious and their percentile group

was assigned at random.









CHAPTER 3
RESULTS

Reading Span Task

Participants who made errors more than 15% in deciding whether a sentence made sense

were excluded from the analyses. We used an existing standard of 85% accuracy criterion for all

participants (Conway et al., 2005). Eleven participants were excluded using this standard.

Additional participants were excluded because they missed one page of the state appraisal scale

or because their native language was not English (i.e., the RSPAN task requires the participants

to be a native English speaker). In sum, we excluded 13 participants' data from the 157

participants. This resulted in removing 8% of the data.

We used working memory total scores in the analyses. Total scores were calculated by

summing the total number of letters recalled in the correct position (First RSPAN total scores: M

= 54.60, SD = 10.78; Second RSPAN total scores: M= 54.92, SD = 11.86). The correlation

between these two tasks was 0.728 (p < 0.01), demonstrating parallel-forms reliability between

the two RSPAN tasks.

Participants were split into three working memory capacity (WMC) groups based on the

distribution of the first RSPAN scores (LWM group's scores ranged from 26 to 50, M= 42.48,

SD = 6.35, n = 50; Medium WM group's scores ranged from 51 to 60, M= 55.91, SD = 2.71, n =

46; HWM group's scores ranged from 61 to 73, M= 65.98, SD = 3.85, n = 48). This

classification resulted in approximately equal numbers of individuals with low, medium, and

high spans in the working memory capacity. Only LWMs and HWMs were included in the

analysis of variance (ANOVA). Participants who were categorized into middle WM group were

dropped in the analyses because they were not of theoretical interest to our specific hypotheses.









We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2

(RSPAN: first vs. second) repeated measures analysis of variance (ANOVA) with the

experimental condition of stress (40th negative feedback treatment group and no feedback control

group) and WMC (high and low) as between-subjects factors and the repeated measures on the

RSPAN scores (first and second) as a within-subjects factor. The three-way interaction was not

significant [F(1, 94) = 0.489, n.s.]. A 2 (WMC: HWMs vs. LWMs) x 2 (RSPAN: first vs.

second) interaction was significant [F(1, 94) = 10.83, p < 0.01]. A similar ANOVA with the

treatment groups revealed a significant WMC x RSPAN interaction [F(1, 42)= 8.549, p < 0.01].

For the control groups we did not obtain a significant WMC x RSPAN interaction effect. We

could conclude there was a significant difference on the first and the second RSPAN scores

between HWM and LWM individuals under the treatment condition, but not under the control

condition. HWM and LWM in the treatment groups performed differently on the first and the

second tasks, whereas the control groups did not.

Planned comparisons derived from our hypotheses showed that for the LWM treatment

group, the RSPAN scores increased 2.90 points out of 75 (SE = 1.39) on the second task

compared to the first task [t(21) = -2.09, p < 0.05]; for the HWM treatment group, the RSPAN

scores decreased 3.59 points out of 75 (SE = 1.74) [t(21) = 2.07, p < 0.05 (one-tailed)]. This

confirmed our hypothesis that HWMs choked in the treatment condition, while LWMs did not.

In addition, we found that for LWM control group, the RSPAN scores increased 3.10 points out

of 75 (SE = 1.52) on the second task compared to the first task [t(27) = -2.05, p = 0.050], which

was quite close to significant. This surprising result showed that LWM in both stress conditions

had a trend to increase their performance, which suggested a statistical regression or a practice









effect existed. In this case, we could not conclude that the increasing performance for LWM

treatment group was due to our manipulation.

State Appraisal Score

The state appraisal score was calculated by summing the scores for each item in the state

appraisal scales after reversing the challenging appraisal items (first state appraisal scores: M=

22.38, SD = 4.64; second state appraisal scores: M= 23.92, SD = 6.33). The higher the state

appraisal scores, the higher the threat appraisal.

We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2 (state

appraisal scores: first vs. second) repeated measures ANOVA, with the state appraisal scores

(first and second) as a within-subjects factor. The three-way interaction was not significant [F(1,

94) = 1.673, n.s.]. A 2 (stress: treatment vs. control) x 2 (state appraisal scores: first vs. second)

interaction was significant [F(1, 94) = 7.88, p < 0.01]. A similar ANOVA with the HWM groups

revealed a significant stress x state appraisal interaction [F(1, 46)= 7.136, p < 0.05]. For the

LWM groups we did not obtain a significant stress x state appraisal interaction effect; there was

also no evidence of any main effects. We could conclude that among the individuals with HWM,

there was a significant difference on the first and the second state appraisal scores between the

treatment and the control conditions, but not among the individuals with LWM.

Planned comparisons derived from our hypotheses showed that for the HWM treatment

group, the state appraisal scores increased 2.86 points out of 48 (SE = 0.98) on the second task

comparing to the first task [t(21) = -2.932, p < 0.01]; for the HWM control group, no difference

was observed [t(21) = 0.738, n.s.]. This confirmed our hypothesis that in the treatment condition

HWMs demonstrated more threat appraisal from first task to the second task.









Self-Reported Appraisal Scale

Following the procedure that Kelsey, Blascovich, Leitten, Schneider, Tomaka, and Wiens

(2000) employed, we summed the scores for the five self-report appraisal items after reversing

some items as needed (first self-reported appraisal scale: M= 19.30, SD = 6.55; second self-

reported appraisal scale: M= 26.78, SD = 7.21). The higher the self-reported appraisal score, the

higher the threat appraisal.

We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2 (self-

reported appraisal scores: first vs. second) repeated measures ANOVA, with the self-reported

appraisal scores (first and second) as a within-subjects factor. The three-way interaction was not

significant [F(1, 94) = 0.656, n.s.]. A 2 (stress: treatment vs. control) x 2 (state appraisal scores:

first vs. second) interaction was significant [F(1, 94) = 8.35, p < 0.01]. A similar ANOVA with

the HWM groups revealed a significant stress x state appraisal interaction [F(1, 46)= 6.75, p <

0.05]. For the LWM groups we did not obtain a significant stress x state appraisal interaction

effect; or evidence of any main effects. We could conclude that among the individuals with

HWM, there was a significant difference on the first and the second self-reported appraisal

scores between the treatment and the control condition, but not among the individuals with

LWM.

Planned comparisons derived from our hypotheses showed that for the HWM treatment

group, the self-reported appraisal scores increased 8.36 points out of 50 (SE = 1.58) on the

second task compared to the first [t(21) = -5.28, p < 0.01]. This confirms our hypothesis that in

the treatment condition, HWMs increased in threat appraisal from the first task to the second

task. For the HWM control group, a similar difference of 3.38 points (SE = 1.15) was also

observed [t(21) = -2.953, p < 0.01]. This was not consistent with our hypothesis, that there would









be no differences in the control groups. However, we could see that HWMs increased much

more under the treatment condition than under the control condition (8.36 comparing to 3.38).

State Anxiety Scale

This scale was calculated by summing the scores for the eight items after reversing some

items as needed (first state anxiety scale: M= 33.62, SD = 8.31; second state anxiety scale: M=

33.80, SD = 9.36). The higher the score, the higher the state anxiety. We did not get any

significant interactions or main effects in the ANOVA analyses (stress x WMC x two repeated

measures on state anxiety). Therefore we could conclude that there were no changes in the

anxiety for either group. This result was not consistent with our hypotheses and findings from

Beilock and Carr's (2005) and Gimming et al.'s (2006) that HWMs became more anxious

whereas LWMs became less anxious under stress.

Stress Score

The self-reported ratings on the question "How stressful was the task you just completed?"

were used to calculate the stress scores (first stress scores: M= 5.91, SD = 2.04; second stress

scores: M= 6.01, SD = 2.35). We did not get any significant interactions or main effects in the

ANOVA analyses (stress x WMC x two repeated measures on stress). Therefore we could

conclude that there were no changes in the stress scores for either group.

Effort Score

The self-reported ratings on the question "How much effort did you put in?" were used to

calculate the effort scores (first effort scores: M = 8.65, SD = 1.34; second effort scores: M=

8.24, SD = 1.77). We did not get any significant interactions in the ANOVA analyses (stress x

WMC x two repeated measures on effort). Therefore we could conclude that there were no

significant changes in the effort score for either group.









Self-Handicapping Scale

Following Rhodewalt's (1990) procedure, we summed the scores for the items after

reversing some items as needed (M= 75.97, SD = 11.69). The lower the score, the lower the self-

handicapping. The self-handicapping scale was only measured once so we conducted 2(stress:

treatment vs. control) x 2(WMC: HWM vs. LWM) ANOVA. However, we did not get any

significant interactions or main effects. Therefore we could conclude that there were no

differences in the self-handicapping scores for either group.

Demographic Questionnaire

We were unable to analyze the effects for gender and ethnicity because of the unequal

distribution in our sample (n for gender: male = 16, female = 128; n for ethnicity: Indian = 1,

Asian = 7, Black = 19, Hispanic =22, White = 89, other = 6).









CHAPTER 4
DISCUSSION

The experiment presented here represents an important departure from traditional choking

research. Customarily, research exploring choking focuses on the negative influence of stress,

typically anxiety (Gimming et al., 2006). Results usually report that only high-powered people

choke, (e.g., individuals with high working memory). Often these results are explained by high-

powered people's "susceptibility to failure", which implies that performance pressure consumes

working memory capacity they rely on to achieve superior performance (Beilock & Carr, 2005).

Previous research has not provided a theory or mechanism to explain why LWMs were less

affected by stress, or, in some cases actually increased in performance. The purpose of the

present experiment was to begin documenting the beneficial as well as detrimental influence of

stress by applying cognitive appraisal theory (Lazarus & Folkman, 1984).

In our experiment, we used social comparison to introduce pressure. We presented a high-

pressure scenario to participants to test whether performance was affected by this challenge (as

choking research indicates), and to explore the reason for performance change. In sum, we

proposed that the difference in pressure (no feedback vs. 40th percentile failure) would either

motivated or hindered people's performance as a function of their working memory capacities.

We expected to find choking in HWMs when given the negative feedback; and LWMs would be

motivated by the negative feedback. We also expected that choking in HWMs was due to

increased anxiety, increased threat appraisal, decreased challenging appraisal, and increased self-

handicapping. In contrast, LWMs' increased performance was due to decreased anxiety,

decreased threat, decreased challenging appraisal, and decreased self-handicapping.

Specifically, we hypothesized that: (1) Negative feedback would result in choking in

HWMs but not in LWMs. Indeed, we found that HWMs choked; in addition, we found LWMs'









performance increased on the second test. Only HWMs in the treatment condition choked, the

control group of HWMs did not show a change in performance. Our results are consistent with

Beilock and Carr's (2005) claim that under high-pressure performance situations, HWM

participants showed decrements in performance. Our results also mirrored Lai's (2008) finding

that while HWM participants did significantly worse after receiving negative feedback, LWM

participants' performance actually improved. It was worth noting that in our experiment, LWMs

in both stress groups had a trend to increase their performance. One possible reason of LWMs'

increasing in scores was a result of statistical regression. That is, scores on the posttest regressed

to the population mean. However, we did not find the statistical regression on HWMs because

only the treatment group demonstrated choking. Alternatively, LWMs' increasing performance

in both conditions might be a result of practice effects. Klein and Fiss (1999) found that scores

on the working memory operation span task markedly increased from the first to the second

administration of the test, indicating a practice effect. In our study, LWMs might have more

room to improve. It is important to note that while LWM participants' performance increased

after receiving the feedback, they still underperformed relative to HWM participants.

We also hypothesized: (2) HWMs would demonstrate more state anxiety than LWMs in

the treatment condition; (4) HWMs would feel more stressful, put less effort and express high

self-handicapping in the second task than in the first task in the treatment condition: We did not

find any changes on these variables in our experiment. In Gimmig et al.'s (2006) study, they

found that anxiety mediating choking. We obtained no evidence that WMC was jeopardized by

pressure via increased anxiety. In Gimmig et al.'s (2006) study, they also found that HWMs had

increased self-handicapping under the high pressure. We did not find this effect. Our inability to

replicate Gimmig's results could be attributed to the fact that our stress manipulation was









different from theirs. Different stressors could result in different anxiety arousal levels. In

Gimming's study, participants were told that the experiment was associated with academic

success, which was defined as high-pressure condition. In our study, we used negative

performance feedback to induce stress.

We hypothesized: (3) HWM individuals would show more threat appraisal in the second

task than in the first task in the treatment condition. The state threat appraisal scores and the self-

reported appraisal scores for HWM under the treatment condition increased, while we did not

find LWMs increased in threat appraisal. Our review of the previous research suggests that

HWMs experienced more pressure than their LWM counterparts (Beilock and Carr, 2005;

Gimmig et al., 2006). This was reflected by highly increased threat appraisal in our study.

However, we also found HWM control group increased the self-reported threat appraisal scores

from the first to the second task but it was a smaller increase compared to the HWM treatment

condition (3.38 points comparing to 8.36 points). Perhaps there is a threshold on the amount of

threat appraisal necessary to decrease performance.

We hypothesized: (5) In the treatment condition, LWMs would show more challenge

appraisal in the second task than in the first task; (6) In the treatment condition, LWMs would

feel less stressful, put more effort and express lower self-handicapping in the second task than in

the first task: We did not find these results in our experiment. LWMs remained the same from

the first to the second task. This result is inconsistent with Lai's (2008) hypotheses that negative

feedback might have acted as a motivator for our participants who had lower working memory

scores. LWMs did not interpret the negative feedback as a motivator as reflected in our measures

of how "challenged" they felt. We are unable to explain LWMs' increasing performance by the

change in their mental state (i.e., appraisal, anxiety).









Lastly, we hypothesized: (7) In the no-feedback condition, the control groups would have

no significant changes in state anxiety, cognitive appraisal, RSPAN performance, stress, effort in

the first and the second task. This was confirmed for all the factors, except that the HWM control

group increased the self-reported threat appraisal scores from the first to the second task. But it

was a relatively small increase compared to the HWM treatment condition.

The self-reported statement questionnaire (Appendix B), which asked how participants felt

after they received the feedback, provided us with open comments. For HWM treatment group,

one theme that emerged was the comments that students were scared by the negative feedback as

so the test was more difficult than they could handle. This is consistent with our results that

HWM treatment group experienced the highest threat appraisal. It is worth noting that some

LWMs in both conditions said that they felt the second test to be easier. Since the test order was

randomized and the mean scores for the two tests were similar (first RSPAN: 54.60; second

RSPAN: 54.92), it is statistically unlikely that one test was inherently more difficult than the

other.

Results from this study broadened the literature on individual differences between span

types. Previous studies usually reported that only HWMs choked. Gimming (2006) reported that

HWMs were more likely to have intrusive thoughts, such as worries, reducing the storage and

processing capacity of working memory, as well as attentional resources, resulting in

performance decrements. In our study, HWMs in the treatment condition felt pressure to

overcome failure thus drawing their attention away from the working memory task. This was

reflected by a significant increase in threat appraisal from the first to the second task. Threat

appraisals are those in which the perception of danger exceeds the perception of abilities or

resources to cope with the stressor, thus decreasing performance (Lazarus & Folkman, 1984).









HWMs appraised the negative feedback as "threatening" and did not have confidence to

overcome. External (i.e., feedback) and internal (i.e., feelings and thoughts) distracters could

draw attentional resources away from a task (Barrett et al., 2004). HWMs' performance

decrement can be explained through the reallocation of attentional resources from task-relevant

information to task-irrelevant worries. On the other hand, we obtained no evidence of choking

for LWMs; probably due to the absence of performance pressure to overcome competency threat,

since LWMs remained in a neutral state (no changes in anxiety or cognitive appraisal from the

first to the second task).

Previous research has suggested that LWMs increase their performance under stressful

situations (Lai, 2008), and also demonstrate lower anxiety comparing to HWMs (Gimming et al.,

2006). However, we did not get this result because LWMs increased their performance

regardless of treatment manipulation. We failed to relate the beneficial influence of stress by

applying cognitive appraisal theory because we could not conclude that the increase of

performance was due to the stress; and there were no changes in cognitive appraisal from the

first to the second task. The increase of performance might be a result of practice effect or

statistical regression. The qualitative data from the statement questionnaire (Appendix B)

showed some LWMs felt the second test to be easier than the first, suggesting that the increasing

performance could be due to practice effects.

While the findings from this study helped to move forward our understanding of working

memory, a few limitations and shortcomings exist. For example, the sample size was small (less

than 30) for each group. Therefore, some of our results did not reach significance (i.e., LWM

control group's increased performance from first to the second task). Moreover, because we used









test-retest, we were unable to separate the confounding effect of stress, practice effect and

statistical regression. Other types of stressors should be considered in future studies as well.

The strength of our present study was to more precisely specify the reason for choking, by

aligning previous work with cognitive appraisal theory. The study explores both the negative and

(potentially) positive effects resulted from stress. Our findings could change the focus of choking

research (i.e., simply anxiety) to a more complex understanding of the appraisal of stress. We

argue that increased threat appraisal in high pressure could result in choking, and provides a new

research direction in this area.

A great amount of effort has been expended examining the effects of feedback given to

students in the education system. The use of negative feedback as stressors in our study helped

contribute to this area. Our findings suggest that some students may interpret negative feedback

counterproductively and the negative feedback might discourage them from achieving.

Specifically, our results point out, assuming that HWM participants are generally high achievers,

those students may interpret negative feedback as a threat, which gives rise to their choking in

stressful situations. This might be helpful for educators and policy makers to determine how and

when to highlight the students' performance in comparison to peers.









APPENDIX A
DEMOGRAPHIC QUESTIONNAIRE

What is your age?

What is your gender?

1. Female
2. Male

How would you classify yourself?

1. American Indian or Native Alaskan
2. Asian or Pacific Islander
3. Black (not of Hispanic origin)
4. Hispanic or Latino
5. White
6. Other:









APPENDIX B
STATEMENT QUESTIONNAIRE

During the second part of the task, do you perceive your performance change in the second

task comparing to the first part? How do you expect your performance in the second part

comparing to the first part? Do you change your effort and motivation in the second task?



Do you become lazy in the second task comparing to the first task? (This question is on the

reverse side of the questionnaire)









APPENDIX C
SCRIPT

The experimenter says to participants, "We are currently collecting data to create a

student normal of University of Florida and we want to know how this test correlates with

academic and cognitive ability. The test is related to academic potential. It would be expected to

serve as a part of standard test in college exam like SAT in future, please take it seriously. The

test is a reliable measure of working memory capacity. Working memory capacity is the ability

to hold different pieces of information simultaneously while trying to process one specific piece

of information. Your score is based on how accurately you evaluate sentences and the number of

letters you can recall in order correctly."









LIST OF REFERENCES


Aronson, J., Lustina, M. J., Good, C., Keough, K., Steele, C. M., & Brown, J. (1999).
When White men can't do math: necessary and sufficient factors in stereotype
threat. Journal ofExperimental Social Psychology, 35, 29-46.

Ashcraft, M. H. (2002). Math anxiety: Personal, educational, and cognitive
consequences. Current Directions in Psychological Science, 11, 182-185.

Ashcraft, M. H., & Kirk, E. P. (2001). The relationship among working memory, math
anxiety, and performance. Journal ofExperimental Psychology: General, 130,
224-237.

Baddeley, A. D. (1990). Human Memory: Theory and Practice. Hove: Lawrence
Associates.

Baddeley, A. D. (2002). Is working memory still working? European Psychologist, 7,
85-97.

Barret, L. F., Tugade, M.M., & Engle, R. W. (2004). Individual differences in working
memory capacity and dual-process theories of the mind. Psychological
Bulletin, 130, 533-573.

Beilock, S. L., & Carr, T. H. (2001). On the fragility of skilled performance: What
governs choking under pressure? Journal of Educational Psychology: General,
130, 701-725.

Beilock, S. L., & Carr, T. H. (2005). When high-powered people fail: Working memory
and "choking under pressure" in math. American Psychological Society, 16,
101-105.

Beilock, S. L., Rydel, R. J., & McConnell, A. R. (2007). Stereotype threat and working
memory: Mechanisms, alleviation, and spillover. Jounral ofExperimental
Psychology: General, 136, 256-276.

Beilock, S. L., & DeCaro, M. S. (2007). From poor performance to success under
stress: Working memory, strategy selection, and mathematical problem
solving under pressure. Journal ofExperimental Psychology: Learnig.
Memory, andCognition, 33, 983-998.

Bosson, J. K., Haymovitz, E. L., & Pinel, E. C. (2004). When saying and doing diverge:
the effects of stereotype threat on self-reported versus non-verbal anxiety.
Journal ofExperimental Social Psychology, 40, 247-255.

Cadinu, M., Maass, A., Frigerio, S., Impagliazzo, L., & Latinotti, S. (2003). Stereotype
threat: The effect of expectancy on performance. European Journal of Social
Psychology, 33, 267-285.









Cheryan, S. C., & Bodenhausen, G. V. (2000). When positive stereotypes threaten
intellectual performance: The psychological hazards of "model minority"
status. American Psychological Society, 11, 399-402.

Croizet, J., & Claire, T. (1998). Extending the concept of stereotype and threat to
social class: the intellectual underperformance of students from low
socioeconomic backgrounds. Personality and Social Psychology Bulletin, 24,
588-594.

Conway, A. R. A., Kane, M. J., Bunting, M. F., Hambrick, D. Z., Wilhelm, O., & Engle, R.
(2005). Working memory span tasks: A methodological review and users
guide. Psychonomic Bulletin and Review, 12(5), 769-786.

Corpus, J., Ogle, C., & Love-Geiger, K. (2006). The effects of social-comparison versus
mastery praise on children's intrinsic motivation. Motivation and Emotion,
30(4), 335-345.

Daneman, M., & Carpenter, P. A. (1980). Individual-differences in working memory
and Reading. Journal of Verbal Learning and Verbal Behavior, 19, 450-466.

Dunning, D., Heath, C., & Suls, J. M. (2004). Flawed self-assessment: Implications for
health, education, and the workplace. Psychological Science in the Public
Interest, 5(3), 69-106.

Engle, R.W., Kane, M. J., & Tuholski, S.W. (1999). Individual differences in working
memorycapacity and what they tell us aboutcontrolled attention, general
fluid intelligence, and function oftheprefrontal cortex. In A. Miyake & P. Shah
(Eds.), Models of working memory: Mechanisms of active maintenance and
executive control (pp. 102-134). New York: Cambridge University Press.

Eysenck, M. W. (1992). Anxiety: The Cognitive Perspective. Hove: Lawrence Erlbaum
Associates.

Folkman, S., Lazarus, R. S., Dunkel-Schetter, C., Delongis, A., & Gruen, R. (1986).
Dynamics of a stressful encounter: Cognitive appraisal, coping, and
encounter outcomes. Journal of Personality and Social Psychology, 50, 992-
1003.

Gaab, J., Rohleder, N., Nater, U. M., & Ehlert, U. (2005). Psychological determinants of
thecortisol stress response: The role of anticipatory cognitive appraisal.
Psychoneuroendocrinology, 30, 599-610.

Gimmig, D., Huguet, P., Caverni, J., & Cury, F. (2006). Choking under pressure and
working memory capacity: When performance pressure reduces fluid
intelligence. Psychonomic Bulletin & Review, 13, 1005-1010.









Kelsey, R. M., Blascovich, J., Leitten, C. L., Schneider, T. R., Tomaka, J., & Wiens, S.
(2000). Cardiovascular reactivity and adaptation to recurrent psychological
stress: The moderating effects of evaluative observation. Psychophysiology,
37, 748-756.

Klein, K., & Fiss, W. H. (1999). The reliability and stability of the Turner and Engle
working memory task. Behavior Research Methods, Instruments, & Computers,
31, 429-432.

Lai., S. (2008). Competency threat: What are the effects on working memory?
Unpublished master's thesis, University of Florida, Gainesville.

Lazarus, R. S., & Folkman, S. (1984). Stress, Appraisal and Coping. New York:
Springer.

Osborne, J. W. (2001). Testing stereotype threat: Does anxiety explain race and sex
differences in achievement? Contemporary Educational Psychology, 26, 291-
310.

Rhodewalt, F. (1990). Self-handicappers: Individual differences in the preference for
anticipatory self-protective acts. In R. L. Higgins, C. R. Snyder, & S. Berglas
(Eds.), Self-handicapping: The paradox that isn't (pp. 69-106). New York:
Plenum.

Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat
reducesworking memory capacity. Journal ofPersonality and Social
Psychology, 85, 440-452.

Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat
reduces working memory capacity. Journal ofPersonality & Social Psychology,
85, 440-452.

Schoofs, D., Preub, D., & Wolf, O. T. (2008). Psychosocial stress induces working
memory impairments in an n-back paradigm. Psychoneuroendocrinology, 33,
643-653.

Skinner, N., & Brewer, N. (2002). The dynamics of threat and challenge appraisals
prior tostressful achievement events. Journal ofPersonality and Social
Psychology, 83, 678-692.

Spencer, S. J., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and women's
math performance. Journal ofExperimental Social Psychology, 35, 4-28.

Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the State-Trait
Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.









Stangor, C., Carr, C., & Kiang, L. (1998). Activating stereotypes undermines task
performance expectations. Journal ofPersonality and Social Psychology, 75,
1191-1197.

Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity
and performance. American Psychologist, 52, 613-629.

Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test
performance of African-Americans. Journal ofPersonality and Social
Psychology, 69, 797-811.

Steele, C. M., Spencer, S., & Aronson, J. (2002). Contending with group image: The
psychology of stereotype and social identity threat. In M. Zanna (Ed.).
Advances in experimental social psychology (Vol. 34). New York: Academic
Press.

Thompson, T., & Dinnel, D. L. (2007). Is self-worth protection best regarded as
intentional self-handicapping behavior or an outcome of choking under
pressure? Educational Psychology, 27, 509-531.

Tomaka, J., Blascovich, J., Kelsey, R. M., & Leitten, C. L. (1993). Subjective,
physiological, and behavioral effects of threat and challenge appraisal.
Journal ofPersonality and Social Psychology, 65, 248-260.

Vick. S. B., Seery, M. D., Blascovich, J., & Weisbuch, M. (2008). The effect of gender
stereotype activation on challenge and threat motivational states. Journal of
Experimental Social Psychology, 44, 624-630.









BIOGRAPHICAL SKETCH

Ye Wang was born and grew up in Shanghai, China. She earned her Bachelor of Arts in

Philosophy from Peking University in China in 2007. In fall 2007 she began her graduate study

and majored in educational psychology at the University of Florida in Gainesville, Florida.





PAGE 1

1 CHOKING UNDER PRESSU RE AND WORKING MEMOR Y: INFLUENCE OF COGNITI VE APPRAISAL By YE WANG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF M ASTER OF ART S IN EDUCATION UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Ye Wang

PAGE 3

3 To Jie Zou

PAGE 4

4 ACKNOWLEDGMENTS I wish to thank several individuals for their help and support concerning my thesis study. First, I thank Dr. David Therriault, my advisor and the chair of my thesis committee, for providing me with generous support throughout the process. His constructive and prompt feedback on my study design, data analyses, and writing was greatly appreciated. I also thank the other member of my thesis commit tee, Dr. David Miller, for his service and helpful comments about my study. Last but not the least, I acknowledge Jie Zou, my husband, for supporting and encouraging me as well as helping me with the writing style. I thank my parents for their love and enc ouragement, and unconditional support for my interests and dreams. All these individuals have helped me achieve a goal that seemed like a very daunting challenge.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 ABSTRACT ................................ ................................ ................................ ................................ ..... 7 CHAPTER 1 OVERVIEW ................................ ................................ ................................ ............................. 8 2 METHODS ................................ ................................ ................................ ............................. 17 Sample ................................ ................................ ................................ ................................ .... 17 Materials ................................ ................................ ................................ ................................ 17 Bell Curve Survey ................................ ................................ ................................ ........... 17 Working Memory Test ................................ ................................ ................................ .... 17 Self Handicapping Scale ................................ ................................ ................................ 19 Demographic Questionnaire ................................ ................................ ............................ 19 State Appraisal Measure ................................ ................................ ................................ .. 19 Self Reported Appraisal and Coping Expectancies ................................ ........................ 20 State Anxiety Scale ................................ ................................ ................................ .......... 21 Procedure ................................ ................................ ................................ ................................ 21 3 RESULTS ................................ ................................ ................................ ............................... 24 Reading Span Task ................................ ................................ ................................ ................. 24 State Appraisal Score ................................ ................................ ................................ .............. 26 Self Reported Appraisal Scale ................................ ................................ ................................ 27 State Anxiety Scale ................................ ................................ ................................ ................. 28 Stress Score ................................ ................................ ................................ ............................. 28 Effort Score ................................ ................................ ................................ ............................. 28 Self Handicapping Scale ................................ ................................ ................................ ........ 29 Demographic Questionnaire ................................ ................................ ................................ ... 29 4 DISCUSSION ................................ ................................ ................................ ......................... 30 APPENDIX A D EMOGRAPHIC Q U ESTIONNAIRE ................................ ................................ .................. 36 B STATEMENT QUESTIONNAIRE ................................ ................................ ....................... 37 C SCRIPT ................................ ................................ ................................ ................................ ... 38 LIST OF REFERENCES ................................ ................................ ................................ ............... 39

PAGE 6

6 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ......... 43

PAGE 7

7 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Art s in Education CHOKING UNDER PRESSURE AND WORKING MEMORY: INFLUENCE OF COGNITIVE APPRAISAL By Ye Wang August 2010 Chair: David Therriault Major: Educational Psychology Eysenck (1992) proposed that an xiety could influence cognition by occupying cognitive processing resources, jeopardizing working memory capacity by pressure via increasing state anxiety. Recent findings (Beilock & Carr, 2001, 2005; Gimming, Huguet, Caverni, & Cury, 2006) have demonstrat ed that only individuals with high working memory (HWM) capacity individuals with low working memory (LWM) capacity increased their performance under such conditions. Gimming and his colleagues suggested that LWM and HWM individuals interpreted stress differently, but research has not explored how these groups may be different. Our present study examined how LWM and HWM individuals interpreted stress differently by applying co gnitive appraisal theory (Lazarus and Folkman, 1984). Our results suggested that HWM group experienced highest threat appraisal and thus choked under pressure.

PAGE 8

8 CHAPTER 1 OVERVIEW Working memory plays an essential role in everyday life; it helps us to kee p track of where we are and what we are doing moment to moment. It aids us in holding information long enough to make a decision. Storing while simultaneously processing information is fundamental to many cognitive tasks such as reasoning, reading comprehe nsion and learning (Baddeley, 2002). Working memory capacity (WMC) is also limited. Individuals with high working memory have greater resources at their disposal to control processing and maintain information. Controlled processing is required for gating, blocking, or suppressing distracting information in the face of distraction and interference in order to maintain temporary goals (Engle, Kane, & Tuholski, 1999). There are many potential influences on working memory. For example, working memory is su sceptible to interference and intrusions. One source of interference and intrusion comes from anxiety, or stress. Eysenck (1992) proposed that anxiety could influence cognition by occupying cognitive processing resources and WMC was jeopardized by pressure via increasing state anxiety (Beilock & Carr, 2001, 2005; Beilock, Tugade, & Engle, 2004). People exposed to stress showed significant working memory impairments (Schoofs, 2008). Ashcraft (2002) found math anxiety disrupted cognitive processing by comprom ising ongoing activity in working memory. Ashcraft and Kirk (2001) found individuals with high math anxiety demonstrated smaller working memory spans. They considered disruption of central executive processes as a possible mechanism and suggested individua l emotional difference variables like math anxiety deserve greater attention, especially on assessments of working memory capacity and functioning. One research approach to induce anxiety and stress in laboratory settings is by using stereotype threat. Ac cording to stereotype threat theory (Steele & Aronson, 1995), performing in

PAGE 9

9 a negatively stereotyped group produces feelings of anxiety, uncertainty, and discomfort. Stereotype threat often impairs performance (Aronson, Lustina, Good, Keough, Steele, & Bro wn, 1999; Croizet & Claire, 1998; Steele & Aronson, 1995). Considerable research has indicated that cognitive and affective consequences of activating a negative self relevant stereotype of race (Cheryan & Bodenhausen, 2000), social economic statues (Steel e, 1997), or gender (Spencer, Steele, & Quinn, 1999; Steele, 1997), including decreased performance expectations (Cadinu, Maass, Frigerio, Impagliazzo, & Latinotti, 2003; Stangor, Carr, & Kiang, 1998), increased anxiety (Bosson, Kaymovitz, & Pinel, 2004; O sborne, 2001), and reduced working memory capacity (Beilock, Rydel, & McConnell, 2007; Schmader & Johns, 2003). tions where the pressure to perform is high (Beilock & Carr, 2001). Having examined the relationship between choking under pressure and individual differences in working memory capacity, some authors concluded that the individuals who were most likely to f ail under performance pressure were those who, in the absence of pressure, had the highest potential for success (i.e., high working memory individuals). Beilock and Carr (2005) found that only high working memory (HWM) individuals were harmed by performan ce pressure in math problems. Gimmig, Huguet, Caverni and Cury (2006) also found that individuals with high working memory were more likely than their low working memory (LWM) counterparts to choke under pressure on working memory and fluid intelligence te given negative feedback about their performance) actually improved on working memory tasks. In contrast, HWM individuals who were given the same feedback demonstrated choking.

PAGE 10

10 Several potential reasons for choking under pressure have been proposed. Anxiety is the one that has been most explored (Beilock & Carr, 2001; Beilock & Carr, 2005; Gimming et al., 2006). Beilock and Carr (2001) explained that heightened anxiety induced se lf focused attention, thus producing intrusive thoughts. Beilock and Carr (2005) explained that performance pressure increased reported state anxiety. Intrusive thoughts, such as worries, reduced the storage and processing capacity of working memory, anxie ty providing situations reduced attentional working memory advantage might make them susceptible to failure when pressure was added, if pressure induced consumption of working memory denied them the capacity they normally relied on to produce their superior performance. In line with Beilock and Carr (2005), Gimming and his colleagues (2006) found anxiety mediated choking among HWM individuals, and choking under pressu re was due to increased state anxiety for HWM individuals. Interestingly, reported decreased anxiety and increased performance. They suggested that LWM and HWM individ uals might interpret high stress situations differently. Anxiety increased in HWM individuals under pressure, but decreased in LWM individuals. They explained that only HWM individuals choked because of their anxious perception of high stakes situations at the onset. In sum, reactions to pressure appear to depend on WMC. It appears that choking only occurs in HWM individuals. Previous studies are less clear about the specific mechanism driving these results. According to Eysenck (1992), anxiety occupies co gnitive processing resources. The question remains, why is this effect only occur for HWM individuals. In fact, LWM individuals demonstrated decreased anxiety (Gimming, et al, 2006) and even increased performance (Lai, 2008) under high pressure situation. Gimming and his colleagues (2006) suggested that the

PAGE 11

11 pressure also affected LWM individuals, but their performance increased in pressure situation. If for LWM indivi duals? The purpose of our study was to further explore the relationship of anxiety and working memory. Specifically, we examined the link between working memory and stress by employing cognitive appraisal theory. In cognitive appraisal theory, stress can h ave differential effects. The effect of stress can be either positive or negative. Negative effect of stress causes anxiety and decreases performance, referred as threat appraisal. Positive effect of stress increases performance on a task causing what is r eferred as challenge appraisal (Lazarus & Folkman, 1984). For example, Vick, Seery, Blascovich and Weisbuch (2008) found an effect of gender stereotype activation related to challenge or threat cognitive appraisal states. Girls, who were told that experime nt was gender biased, would feel gender stereotype threat; therefore they exhibited the threat state and underperformed. If girls were told that experiment was gender fair, they would not feel stereotype threat; instead, they exhibited challenge appraisal and perform better. In contrast, if the experimenter told boys that the experiment was gender fair, they showed threat and underperformed. If boys were told that experiment was gender biased (i.e., females would underperform), they would show the challenge state and increase performance. Thus, studies regarding cognitive appraisal theory suggested that interpretation of stress could either increase or decrease performance, depending on the appraisal of the stress. In cognitive appraisal theory, Lazarus and Folkman (1984) proposed two major components of the processes: primary appraisal and secondary appraisal. Primary appraisal reflects perceptions of the nature and degree of risk in the situation. Secondary appraisal reflects perceptions of resources or ab ilities to cope with the situation. According to appraisal theories,

PAGE 12

12 ither exceeds the perception of abilities or resources to cope with the stressor, and result in underperformance. Challenge appraisals, in contrast, are those in which the perception of danger does not exceed the abilities to cope, and possibly increase performance. In a serial subtraction tasks (i.e., the participant was asked to repeatedly subtract a small number, like 7 or 13, from a running total that starts as a four digit number, such as 5,342), threatening appraisals leaded to performances with more mistakes and about a 25% slower rate than the challenging appraisal conditions (Tomaka, Blascovich, Kelsey, & Leitten, 1993). Threat appraisals are more strongly associated with negative emotional reactions, high rates of cognitive interference, and poor performance levels than challenge appraisals are. Threatened individuals may be less task focused and have intrusive cognitions, whereas challenged individuals ma y be less distracted by negative emotions and more task focused. Folkman and his colleagues (1986) also found that participants who appraised self reported stressful events as changeable (high secondary appraisal of coping efficacy) used task focused copin g strategies (e.g., planed problem solving), and were more motivated and more satisfied with encounter outcomes than participants who appraised the events as unchangeable. These participants (low secondary appraisal of coping efficacy) used more emotion fo cused coping strategies (e.g., distancing and escape avoidance). These emotional coping strategies left participants less motivated to finish the task, and more likely to try to find self handicapping excuses. Skinner and Brewer (2002) demonstrated that ch allenge appraisals were associated with more confident coping expectancies, lower perceptions of threat, higher positive emotion, and

PAGE 13

13 more beneficial perceptions of the effects of appraisal and emotion on performance. Generally, higher performance was ass ociated with challenge appraisals. Gaab, Rohleder, Nater, and Ehlert (2005) linked appraisal with psychobiological stress change and found that appraisal helped to predict performance. Vick and his colleagues (2008), as we reviewed above, found an effect o f gender stereotype activation due to challenge/threat states. Females in the gender biased condition exhibited more threatened state, whereas women in the gender fair condition exhibited more challenge state than they did in the gender biased condition. T hus, challenge and threat states can be activated by anxiety of negative information of oneself and change performance. Threatened individuals may be less task focused, and more self emotion focused, whereas challenged individuals may be less distracted by negative emotions, and more task focused. As reviewed above, in working memory research, Beilock and Carr (2001) found self consciousness training, which trained participants not to focus on thoughts about self but focus on task, eliminated choking. Intru sive thoughts, which are less task focused, can cause choking (Beilock and Carr, 2005). Heightened anxiety induces self focused attention and causes choking. Given our review of the literature, it was reasonable to hypothesize that choking is related to co gnitive appraisal. Cognitive appraisal might help to explain why choking works differently in high or low working memory individuals. It is possible that HWM people are more likely to demonstrate threat appraisal and choke under pressure; whereas LWM peopl e are more likely to demonstrate challenge appraisal and avoid choking. Furthermore, because individuals with threat appraisal lose the confidence to cope with the problems they will be more likely to try to employ excuses to avoid the stress, resulting i n high self handicapping scores. In contrast, for challenge appraisal individuals, they are likely to be more motivated and demonstrate low self handicapping scores under stressful situation.

PAGE 14

14 Gimming et al. (2006) found differences in the self handicapping measurements for HWMs and LWMs. Choking was not only related to state anxiety, but also self handicapping. Self handicapping was used to describe the tendency of trying not to focus on task in order to use this as an excuse to avoid pressure. Gimming et a l. (2006) reported that LWM people had less anxiety score on state anxiety scale, and low self (1990), explaining why LWMs did not choke under pressure. In contrast, HWM individuals displayed a tendency to excuse the ir poor performance under pressure (they scored higher in self handicapping than LWM individuals). Thompson (2007) found that poor performance feedback could produce self handicapping behavior in people with high ability, which is analogous to the HWM indi viduals in our study. In our study, we predicted that for LWMs, challenge appraisal would be linked to lower self handicapping scores; and for HWMs, threat appraisal would be linked to higher scores of self handicapping. Therefore, we incorporated in our s tudy a self scale, 1990) at the end of second task and Gimming et al. used the same scale in their study. explored why LWM individuals p erform better and seem to be under less anxiety when placed in high cognitive resources under low stressful task, saving cognitive resources for more stressful tas ks. In our study, we would also explore this hypothesis by giving a self report effort scale at the end of each task. On the other hand, this self report effort scale could also give us important handicapping behavior would cause a withdrawal of effort under high pressure.

PAGE 15

15 The current study extends previous research by examining a potential cognitive appraisal consequence in HWMs and LWMs. The current study was a 2 (WMC: HWMs vs. LWMs) x 2 (stress: treatment vs. control ) design. to induce stress. Participants were categorized into high, middle or low working memory span (while dropping the middle span group in the paired comparison analyses because it w as not of interest to our hypotheses). Participants were administered a working memory test, then given feedback on how well they did by assigning them randomly into one of two groups/conditions: 40 th percentile, or no feedback condition. As noted by Beilo ck and Carr (2005), pressure that caused people to choke included the consequences of suboptimal performance, especially on examinations, including poor evaluations by mentors, teachers and peers. The 40 th percentile condition was expected to create stress for the second task (i.e., the participants were informed that they received a failing score of 60%). Then participants then completed a second working memory task. Participants in the no feedback condition continued on to the second test without any expl icit feedback. We predicted that people with HWM would be more likely to interpret feedback as threat appraisal and choke under pressure while LWM people would not. Our specific hypotheses regarding individuals in the 40 th percentile feedback treatment co ndition were as follows: (1) HWM individuals, not in LWM ones; (2) HWM people would demonstrate more state anxiety than LWM people; This hypothesis could potenti ally provide confirmatory evidence for Beilock (3) HWM individuals would show more threat appraisal in the second task than in the first task; (4) HWMs would fe el more stressful, put less effort and express high self handicapping in the

PAGE 16

16 second task than in the first task; (5) LWMs would show more challenge appraisal in the second task than in the first task; (6) LWMs would feel less stressful, put more effort and express lower self handicapping in the second task than in the first task. In no feedback condition, we hypothesized: (7) the control groups would have no significant changes in state anxiety, cognitive appraisal, RSPAN performance, stress, effort in firs t and second task.

PAGE 17

17 CHAPTER 2 METHODS Sample There were 157 participants in this study. They were undergraduate educational psychology students from the University of Florida who participated for credits toward a course requirement. Materials Bell Curve Survey Upon signing the informed consent, participants were ad ministered a Bell Curve Survey asking them to indicate where they believed their academic competency (percentile) fell on the curve comparing to their peers. Because in our experiment, particip ants were informed that this experiment correlated well with academic competency, the survey served to set the stage for the manipulation of giving negative feedback treatment by highlighting social comparisons between individuals. Working Memory Test We e mployed a dual processing test called the reading span task (RSPAN) to measure working memory capacity (Conway, Kane, Bunting, Hambrick, Wilhelm, & Engle, 2005). The idea is that high working memory capacity individuals do better job in dual tasks because they have more resources to maintain and keep track of multiple sources of information. Daneman and Carpenter (1980) originally developed the RSPAN task to assess working memory during reading. In our study, participants evaluated sentences while memorizin g letters for later recall so it required dual processing. Participants were asked to read a sentence presented on the computer screen, and decided whether or not that sentence made sense in a real world scenario by clicking the screen. A letter that required participants to recall later was

PAGE 18

18 presented for 2 seconds after each sentence decision and a blank screen lasting 1 second separated the presentation of each sentence and letter. At the end of a series of sentence/letter c ombination trials, participants were asked to recall, as best as they could, the letters in the order seconds. The computer recorded the number of letters recalled in the correct order, and or incorrect sentences decision One set ranged from three to six sentences, so at the end of a set participants were asked to recall the letters he/she could recall in the set, and then came to the next set. I n our study, the RSPAN task included 15 sets of sentence/letter trials that contained three to six sentence/letter combinations. The sets were presented in random order so participants could not anticipate how many sentences and letters they were required to evaluate and recall at the beginning of each set. This was a dual task because participants made judgments whether the sentence made sense, and at the same time remembered the letters presented after each sentence. Working memory capacity was scored by the total number of letters that participants recalled correctly from each set while maintaining at least an 85% accuracy rate on judging the sensibility of sentences. The second RSPAN task mirrored the first. Both tasks had equal number of sets and equal number of nonsense and sense sentences. The letters to be recalled were also taken from the same sampling of letters as the first RSPAN task. The sentences from the second RSPAN were of equal length to the first task and same number of words per sentence. A nd the sentences are also of similar difficulty levels using the Flesch Kincaid Grade Level readability scale. We chose RSPAN task because it was stressful and difficult, as research reports had indicated (Conway et al., 2005; Lai, 2008). Stress is most l ikely to occur when the task is

PAGE 19

19 Therefore, this task is likely to induce threat and cognitive appraisal of stress. Self Handicapping Scale Participants were adm inistered a self handicapping scale after completing the second task. They rated statements about how handicapped they had felt during the task performance on a 5 adapted from Rhodewalt, 1990). This served to index self handicapping ( a =0.69). Demographic Questionnaire Participants were administered a brief Demog raphic Questionnaire (Appendix A ) at the end of the study. This was included to insure that there was no accidental priming of a gender or race stereotype threat that could affect performance. The questionnaire asked questions regarding age, gender, and ethnicity. The questionnaire was included as an additional control measure. State Appraisal Measure d an example of this scale used. Four threat ot be able to achieve positive benefits of state appraisals were measured on separate 6 point response scales (1 = strongly d isagree, 6 =

PAGE 20

20 strongly agree and 1 = this thought hardly ever occurred, 6 = this thought occurred almost constantly). Self Reported Appraisal and Coping Expectancies This scale was similar to state appraisal scale because they both measure cognitive apprai sals. However, the self manipulation) because it asked participants about the upcoming task. In contrast, the state appraisal scale is a more global assessment of general state rather tha about the specific task. Tomaka et al. (1993) assessed primary and second appraisal by asking the task. Skinner and Brewer (2002) also measured coping expectancies by self reported anticipated coping ability and expected level of performance. Coping expectancies were assessed by two 10 point scales tapping anticipated coping ability (1 = very confident can cope effectively, 10 = very concerned whether can cope effectively) and the expected level of performanc e (1 = poor, 10 = high quality). The second coping item was reverse coded. Kelsey et think the Obviously (a) was related to threat appraisal; (b) was related to challenge appraisal; (c) was the primary appraisal; (d) and (e) were like what Skinner and Brewer used as secondary appraisal coping expectancy.

PAGE 21

21 point scales for each que stion. These appraisal items were presented to subjects through computer after they had State Anx iety Scale Following each task, participants completed an anxiety scale that has several items related to perceptions of the test and testing situation. Anxiety during the tests was measured using the questions adapted from the Spielberger State Anxiety Sc ale (1970) (see Schmader & Johns, 2003). Participants rated on a 7 point scale (ranging from 1 not at all to 7 very much) how much they felt anxious, comfortable, jittery, worried, at ease, nervous, relaxed, and calm while taking the test. The items on the scale were summed (after reverse scoring comfortable, at ease, relaxed, and calm) to form an index of state anxiety, where higher numbers indicated more anxiety ( a = 0.90). Procedure After seating a participant by a computer in a quiet laboratory room, a research assistant first asked the participant to read and sign an informed consent form. The research assistant then presented the Bell Curve Survey in which part icipants indicated their academic competency in percentile compared to their peers on the bell curve. Participants were told that the experiment consisted two tests and took about 1 hour long. The research assistant highlighted the importance and credibili ty of the test by saying that the test was created by an authority and was related to the academic performance. The research assistant also described working memory capacity so that the participants were made fully aware of the task. The detailed script wa s included in Appendix C The first RSPAN test instruction was then presented on the computer. Immediately after viewing and fully understanding the instruction, participants were measured on the state

PAGE 22

22 appraisal measure and self report appraisal and coping expectancies. Then the first RSPAN test was administered. point scale ), and the stat e anxiety scale. After completing the questions, the research assistant told them their assigned into one of two conditions/groups, which was no feedback, or negative (stress) compared to others. Even though our participants should have had also familiar with this type of reporting which they experience d in standardized tests like the SAT the research assistant still explained the percentile rankings and informed participants how their scores were compared to others. Participants in the 40 th percentile were told that based on their rank they were expected to have difficulty o n the subsequent test. The 40 th percentile was used because it was considered below average so as to create stress, but it was still a realistic achievement group. In the no feedback condition, participants completed the questionnaires and questions but co ntinued onto the second test without any score or expectation feedback Using t he no feedback condition as the control group we would learn whether the two different RSPAN tasks were of equal difficulty level. The second RSPAN test was then administered. After the computer displayed the instructions, participants were assessed on the same self report appraisal and state appraisal scales. Then the second RSPAN test started. After finishing the second RSPAN test, participants were assessed on the same questi ons and state anxiety scales as they did in the first RSPAN test. Upon completing above tests and scales for all participants, the experimenter let participants fill

PAGE 23

23 out the self handicapping scale, and a questionnaire asking how they felt after they recei ved the feedback, and how they perceived their performance, expectancy, effort and motivation changed in the second task (Appendix B ). Participants were then asked to fill out the demographic questionnaire. Finally, participants in the treatment condition were debriefed by informing participants that their percentile rankings feedbacks for the first test were fictitious and their percentile group was assigned at random.

PAGE 24

24 CHAPTER 3 RESULTS Reading Span Task Participants who made errors more than 15% in deci ding whether a sentence made sense were excluded from the analyses. We used an existing standard of 85% accuracy criterion for all participants (Conway et al., 2005). Eleven participants were excluded using this standard. Additional participants were exclu ded because they missed one page of the state appraisal scale or because their native language was not English (i.e., the RSPAN task requires the participants participa nts. This resulted in removing 8% of the data. We used working memory total scores in the analyses. Total scores were calculated by summing the total number of letters recalled in the correct position (First RSPAN total scores: M = 54.60, SD = 10.78; Seco nd RSPAN total scores: M = 54.92, SD = 11.86). The correlation between these two tasks was 0.728 ( p < 0.01), demonstrating parallel forms reliability between the two RSPAN tasks. Participants were split into three working memory capacity (WMC) groups base d on the M = 42.48, SD = 6.35, n M = 55.91, SD = 2.71, n = M = 65.98, SD = 3. 85, n = 48). This classification resulted in approximately equal numbers of individuals with low, medium, and high spans in the working memory capacity. Only LWMs and HWMs were included in the analysis of variance (ANOVA). Participants who were categorized into middle WM group were dropped in the analyses because they were not of theoretical interest to our specific hypotheses

PAGE 25

25 We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2 (RSPAN: first vs. second) repeated measures analysis o f variance (ANOVA) with the experimental condition of stress (40 th negative feedback treatment group and no feedback control group) and WMC (high and low) as between subjects factors and the repeated measures on the RSPAN scores (first and second) as a wit hin subjects factor. The three way interaction was not significant [ F (1, 94) = 0.489, n.s.]. A 2 (WMC: HWMs vs. LWMs) x 2 (RSPAN: first vs. second) interaction was significant [ F (1, 94) = 10.83, p < 0.01]. A similar ANOVA with the treatment groups revealed a significant WMC x RSPAN interaction [ F (1, 42) = 8.549, p < 0.01]. For the control groups we did not obtain a significant WMC x RSPAN interaction effect. We could conclude there was a significant difference on the first and the second RSPAN scores betwee n HWM and LWM individuals under the treatment condition, but not under the control condition. HWM and LWM in the treatment groups performed differently on the first and the second tasks, whereas the control groups did not. Planned comparisons derived from our hypotheses showed that for the LWM treatment group, the RSPAN scores increased 2.90 points out of 75 ( SE = 1.39) on the second task compared to the first task [ t (21) = 2.09, p < 0.05]; for the HWM treatment group, the RSPAN scores decreased 3.59 poin ts out of 75 ( SE = 1.74) [ t (21) = 2.07, p < 0.05 (one tailed)]. This confirmed our hypothesis that HWMs choked in the treatment condition, while LWMs did not. In addition, we found that for LWM control group, the RSPAN scores increased 3.10 points out of 7 5 ( SE = 1.52) on the second task compared to the first task [ t (27) = 2.05, p = 0.050], which was quite close to significant. This surprising result showed that LWM in both stress conditions had a trend to increase their performance, which suggested a stat istical regression or a practice

PAGE 26

26 effect existed. In this case, we could not conclude that the increasing performance for LWM treatment group was due to our manipulation. State Appraisal Score The s tate appraisal score was calculated by summing the scores for each item in the state appraisal scales after reversing the challenging appraisal items (first state appraisal scores: M = 22.38, SD = 4.64; second state appraisal scores: M = 23.92, SD = 6.33). The higher the state appraisal scores, the higher the thr eat appraisal. We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2 (state appraisal scores: first vs. second) repeated measures ANOVA, with the state appraisal scores (first and second) as a within subjects factor. The three way in teraction was not significant [ F (1, 94) = 1.673, n.s.]. A 2 (stress: treatment vs. control) x 2 (state appraisal scores: first vs. second) interaction was significant [ F (1, 94) = 7.88, p < 0.01]. A similar ANOVA with the HWM groups revealed a significant s tress x state appraisal interaction [ F (1, 46) = 7.136, p < 0.05]. For the LWM groups we did not obtain a significant stress x state appraisal interaction effect; there was also no evidence of any main effects. We could conclude that among the individuals w ith HWM, there was a significant difference on the first and the second state appraisal scores between the treatment and the control conditions, but not among the individuals with LWM. Planned comparisons derived from our hypotheses showed that for the HW M treatment group, the state appraisal scores increased 2.86 points out of 48 ( SE = 0.98) on the second task comparing to the first task [ t (21) = 2.932, p < 0.01]; for the HWM control group, no difference was observed [ t (21) = 0.738, n.s.]. This confirmed our hypothesis that in the treatment condition HWMs demonstrated more threat appraisal from first task to the second task.

PAGE 27

27 Self Reported Appraisal Scale Following the procedure that Kelsey, Blascovich, Leitten, Schneider, Tomaka, and Wiens (2000) employed we summed the scores for the five self report appraisal items after reversing some items as needed (first self reported appraisal scale: M = 19.30, SD = 6.55; second self reported appraisal scale: M = 26.78, SD = 7.21). The higher the self reported appra isal score, the higher the threat appraisal. We conducted a 2 (stress: treatment vs. control) x 2 (WMC: HWM vs. LWM) x 2 (self reported appraisal scores: first vs. second) repeated measures ANOVA, with the self reported appraisal scores (first and second) as a within subjects factor. The three way interaction was not significant [ F (1, 94) = 0.656, n.s.]. A 2 (stress: treatment vs. control) x 2 (state appraisal scores: first vs. second) interaction was significant [ F (1, 94) = 8.35, p < 0.01]. A similar ANOV A with the HWM groups revealed a significant stress x state appraisal interaction [ F (1, 46) = 6.75, p < 0.05]. For the LWM groups we did not obtain a significant stress x state appraisal interaction effect; or evidence of any main effects. We could conclud e that among the individuals with HWM, there was a significant difference on the first and the second self reported appraisal scores between the treatment and the control condition, but not among the individuals with LWM. Planned comparisons derived from our hypotheses showed that for the HWM treatment group, the self reported appraisal scores increased 8.36 points out of 50 ( SE = 1.58) on the second task compared to the first [ t (21) = 5.28, p < 0.01]. This confirms our hypothesis that in the treatment co ndition, HWMs increased in threat appraisal from the first task to the second task. For the HWM control group, a similar difference of 3.38 points ( SE = 1.15) was also observed [ t (21) = 2.953, p < 0.01]. This was not consistent with our hypothesis, that t here would

PAGE 28

28 be no differences in the control groups. However, we could see that HWMs increased much more under the treatment condition than under the control condition (8.36 comparing to 3.38). State Anxiety Scale This scale was calculated by summing the scores for the eight items after reversing some items as needed (first state anxiety scale: M = 33.62, SD = 8.31; second state anxiety scale: M = 33.80, SD = 9.36). The higher the score, the higher the state anxiety. We did not get any significant interac tions or main effects in the ANOVA analyses (stress x WMC x two repeated measures on state anxiety). Therefore we could conclude that there were no changes in the anxiety for either group. This result was not consistent with our hypotheses and findings fro m whereas LWMs became less anxious under stress. Stress Score The self were used to calcul ate the stress scores (first stress scores: M = 5.91, SD = 2.04; second stress scores: M = 6.01, SD = 2.35). We did not get any significant interactions or main effects in the ANOVA analyses (stress x WMC x two repeated measures on stress). Therefore we co uld conclude that there were no changes in the stress scores for either group. Effort Score The self calculate the effort scores (first effort scores: M = 8.65, SD = 1.34; seco nd effort scores: M = 8.24, SD = 1.77). We did not get any significant interactions in the ANOVA analyses (stress x WMC x two repeated measures on effort). Therefore we could conclude that there were no significant changes in the effort score for either gr oup.

PAGE 29

29 Self Handicapping Scale reversing some items as needed ( M = 75.97, SD = 11.69). The lower the score, the lower the self handicapping. The self handicapping scale was onl y measured once so we conducted 2(stress: treatment vs. control) x 2(WMC: HWM vs. LWM) ANOVA. However, we did not get any significant interactions or main effects. Therefore we could conclude that there were no differences in the self handicapping scores f or either group. Demographic Questionnaire We were unable to analyze the effects for gender and ethnicity because of the unequal distribution in our sample ( n for gender: male = 16, female = 128; n for ethnicity: Indian = 1, Asian = 7, Black = 19, Hispanic =22, White = 89, other = 6).

PAGE 30

30 CHAPTER 4 DISCUSSION The experiment presented here represents an important departure from traditional choking research. Customarily, research exploring choking focuses on the negative influence of stress, typically anxiety (G imming et al., 2006). Results usually report that only high powered people choke, (e.g., individuals with high working memory). Often these results are explained by high consumes working memory capacity they rely on to achieve superior performance (Beilock & Carr, 2005). Previous research has not provided a theory or mechanism to explain why LWMs were less affected by stress, or, in some cases actually increased in perform ance. The purpose of the present experiment was to begin documenting the beneficial as well as detrimental influence of stress by applying cognitive appraisal theory (Lazarus & Folkman, 1984). In our experiment, we used social comparison to introduce pres sure. We presented a high pressure scenario to participants to test whether performance was affected by this challenge (as choking research indicates), and to explore the reason for performance change. In sum, we proposed that the difference in pressure (n o feedback vs. 40 th percentile failure) would either motivated or hindered people's performance as a function of their working memory capacities. We expected to find choking in HWMs when given the negative feedback; and LWMs would be motivated by the negat ive feedback. We also expected that choking in HWMs was due to increased anxiety, increased threat appraisal, decreased challenging appraisal, and increased self handicapping. In contrast, LWMs' increased performance was due to decreased anxiety, decreased threat, decreased challenging appraisal, and decreased self handicapping. Specifically, we hypothesized that: (1) Negative feedback would result in choking in

PAGE 31

31 performan ce increased on the second test. Only HWMs in the treatment condition choked, the control group of HWMs did not show a change in performance. Our results are consistent with Beilock and Carr's (2005) claim that under high pressure performance situations, H WM participants showed decrements in performance. Our results also mirrored Lai's (2008) finding that while HWM participants did significantly worse after receiving negative feedback, LWM t in our experiment, LWMs increasing in scores was a result of statistical regression. That is, scores on the posttest regressed to the population mean. However, we did not find the statistical regression on HWMs because in both conditions might be a result of practice effects. Klein and Fiss (1999) found that scores on the w orking memory operation span task markedly increased from the first to the second administration of the test, indicating a practice effect. In our study, LWMs might have more increased after receiving the feedback, they still underperformed relative to HWM participants. We also hypothesized: (2) HWMs would demonstrate more state anxiety than LWMs in the treatment condition; (4) HWMs would feel more stressful, put less effort and express high self handicapping in the second task than in the first task in the treatment condition: We did not find any changes on these variables in our experiment. In Gimmig et al.'s (2006) study, they found that anxiety mediating choking. We obtain ed no evidence that WMC was jeopardized by pressure via increased anxiety. In Gimmig et al.'s (2006) study, they also found that HWMs had increased self handicapping under the high pressure. We did not find this effect. Our inability to results could be attributed to the fact that our stress manipulation was

PAGE 32

32 different from theirs. Different stressors could result in different anxiety arousal levels. In c success, which was defined as high pressure condition. In our study, we used negative performance feedback to induce stress. We hypothesized: (3) HWM individuals would show more threat appraisal in the second task than in the first task in the treatment condition. The state threat appraisal scores and the self reported appraisal scores for HWM under the treatment condition increased, while we did not find LWMs increased in threat appraisal. Our review of the previous research suggests that HWMs experienc ed more pressure than their LWM counterparts (Beilock and Carr, 2005; Gimmig et al., 2006). This was reflected by highly increased threat appraisal in our study. However, we also found HWM control group increased the self reported threat appraisal scores f rom the first to the second task but it was a smaller increase compared to the HWM treatment condition (3.38 points comparing to 8.36 points). Perhaps there is a threshold on the amount of threat appraisal necessary to decrease performance. We hypothesize d: (5) In the treatment condition, LWMs would show more challenge appraisal in the second task than in the first task; (6) In the treatment condition, LWMs would feel less stressful, put more effort and express lower self handicapping in the second task th an in the first task: We did not find these results in our experiment. LWMs remained the same from feedback might have acted as a motivator for our partici pants who had lower working memory scores. LWMs did not interpret the negative feedback as a motivator as reflected in our measures change in their mental state (i. e., appraisal, anxiety).

PAGE 33

33 Lastly, we hypothesized: (7) In the no feedback condition, the control groups would have no significant changes in state anxiety, cognitive appraisal, RSPAN performance, stress, effort in the first and the second task. This was c onfirmed for all the factors, except that the HWM control group increased the self reported threat appraisal scores from the first to the second task. But it was a relatively small increase compared to the HWM treatment condition. The self reported statem ent questionnaire (Appendix B ), which asked how participants felt after they received the feedback, provided us with open comments. For HWM treatment group, one theme that emerged was the comments that students were scared by the negative feedback as so th e test was more difficult than they could handle. This is consistent with our results that HWM treatment group experienced the highest threat appraisal. It is worth noting that some LWMs in both conditions said that they felt the second test to be easier. Since the test order was randomized and the mean scores for the two tests were similar (first RSPAN: 54.60; second RSPAN: 54.92), it is statistically unlikely that one test was inherently more difficult than the other. Results from this study broadened th e literature on individual differences between span types. Previous studies usually reported that only HWMs choked. Gimming (2006) reported that HWMs were more likely to have intrusive thoughts, such as worries, reducing the storage and processing capacity of working memory, as well as attentional resources, resulting in performance decrements. In our study, HWMs in the treatment condition felt pressure to overcome failure thus drawing their attention away from the working memory task. This was reflected by a significant increase in threat appraisal from the first to the second task. Threat appraisals are those in which the perception of danger exceeds the perception of abilities or resources to cope with the stressor, thus decreasing performance (Lazarus & Folkman, 1984).

PAGE 34

34 overcome. External (i.e., feedback) and internal (i.e., feelings and thoughts) distracters could draw attentional resources away from a task (Barrett et al decrement can be explained through the reallocation of attentional resources from task relevant information to task irrelevant worries. On the other hand, we obtained no evidence of choking for LWMs; probably due to the absence of performance pressure to overcome competency threat, since LWMs remained in a neutral state (no changes in anxiety or cognitive appraisal from the first to the second task). Previous research has suggested that LWMs increase their performance under stre ssful situations (Lai, 2008), and also demonstrate lower anxiety comparing to HWMs (Gimming et al., 2006). However, we did not get this result because LWMs increased their performance regardless of treatment manipulation. We failed to relate the beneficial influence of stress by applying cognitive appraisal theory because we could not conclude that the increase of performance was due to the stress; and there were no changes in cognitive appraisal from the first to the second task. The increase of performanc e might be a result of practice effect or statistical regression. The qualitative data from the statement questionnaire (Appendix B ) showed some LWMs felt the second test to be easier than the first, suggesting that the increasing performance could be due to practice effects. While the findings from this study helped to move forward our understanding of working memory, a few limitations and shortcomings exist. For example, the sample size was small (less than 30) for each group. Therefore, some of our resu lts did not reach significance (i.e., LWM

PAGE 35

35 test retest, we were unable to separate the confounding effect of stress, practice effect and statistical regression. Other types of stressors should be considered in future studies as well. The strength of our present study was to more precisely specify the reason for choking, by aligning previous work with cognitive appraisal theory. The study explores both the negative and (potentially) positive effects resulted from stress. Our findings could change the focus of choking research (i.e., simply anxiety) to a more complex understanding of the appraisal of stress. We argue that increased threat appraisal in high pressure c ould result in choking, and provides a new research direction in this area. A great amount of effort has been expended examining the effects of feedback given to students in the education system. The use of negative feedback as stressors in our study help ed contribute to this area. Our findings suggest that some students may interpret negative feedback counterproductively and the negative feedback might discourage them from achieving. Specifically, our results point out, assuming that HWM participants are generally high achievers, those students may interpret negative feedback as a threat, which gives rise to their choking in stressful situations. This might be helpful for educators and policy makers to determine how and ormance in comparison to peers.

PAGE 36

36 APPENDIX A DEMOGRAPHIC QUESTION NAIRE What is your age? _____ What is your gender? 1. Female 2. Male How would you classify yourself? 1. American Indian or Native Alaskan 2. Asian or Pacific Islander 3. Black (not of Hispanic origin) 4. Hispanic or Latino 5. White 6. Other: __________________

PAGE 37

37 APPENDIX B STATEMENT QUESTIONNA IRE During the second part of the task, do you perceive your performance change in the second task comparing to the first part? How do you expect your performance in the s econd part comparing to the first part? Do you change your effort and motivation in the second task? Do you become lazy in the second task comparing to the first task? (This question is on the reverse side of the questionnaire)

PAGE 38

38 APPENDIX C SCRIPT The exp student normal of University of Florida and we want to know how this test correlates with academic and cognitive ability. The test is related to academic potential. It would be e xpected to serve as a part of standard test in college exam like SAT in future, please take it seriously. The test is a reliable measure of working memory capacity. Working memory capacity is the ability to hold different pieces of information simultaneous ly while trying to process one specific piece of information. Your score is based on how accurately you evaluate sentences and the number of

PAGE 39

39 LIST OF REFERENCES Aronson, J., Lustina, M. J., Good, C., Keough, K., Steele, C. M., & Brown, J. (1999). threat. Journal of Experimental Social Psychology, 35 29 46. Ashcraft, M. H. (2002). Math anxiety: Personal, educational, and cognitive cons equences. Current Directions in Psychological Science, 11 182 185. Ashcraft, M. H., & Kirk, E. P. (2001). The relationship among working memory, math anxiety, and performance. Journal of Experimental Psychology: General, 130, 224 237. Baddeley, A. D. (1 990). Human Memory: Theory and Practice Hove: Lawrence Associates. Baddeley, A. D. (2002). Is working memory still working? European Psychologist, 7, 85 97. Barret, L. F., Tugade, M.M., & Engle, R. W. (2004). Individual differences in working memory capacity and dual process theories of the mind. Psychological Bulletin, 130, 533 573. Beilock, S. L., & Carr, T. H. (2001). On the fragility of skilled performance: What governs choking under pressure? Journal of Educational Psychology: General, 130, 70 1 725. Beilock, S. L., & Carr, T. H. (2005). When high powered people fail: Working memory American Psychological Society, 16, 101 105. Beilock, S. L., Rydel, R. J., & McConnell, A. R. (2007). Stereotype threat and working memory: Mechanisms, alleviation, and spillover. Jounral of Experimental Psychology: General, 136 256 276. Beilock, S. L., & DeCaro, M. S. (2007). From poor performance to success under stress: Working memory, strategy selection, and mathematica l problem solving under pressure. Journal of Experimental Psychology: Learning, Memory, andCognition, 33, 983 998. Bosson, J. K., Haymovitz, E. L., & Pinel, E. C. (2004). When saying and doing diverge: the effects of stereotype threat on self reported v ersus non verbal anxiety. Journal of Experimental Social Psychology, 40, 247 255. Cadinu, M., Maass, A., Frigerio, S., Impagliazzo, L., & Latinotti, S. (2003). Stereotype threat: The effect of expectancy on performance. European Journal of Social Psych ology, 33, 267 285.

PAGE 40

40 Cheryan, S. C., & Bodenhausen, G. V. (2000). When positive stereotypes threaten status. American Psychological Society, 11, 399 402. Croizet, J., & Claire, T. (19 98). Extending the concept of stereotype and threat to social class: the intellectual underperformance of students from low socioeconomic backgrounds. Personality and Social Psychology Bulletin, 24 588 594. Conway, A. R. A., Kane, M. J., Bunting, M. F. Hambrick, D. Z., Wilhelm, O., & Engle, R. (2005). Working memory span tasks: A methodological review and users guide. Psychonomic Bulletin and Review, 12(5), 769 786. Corpus, J., Ogle, C., & Love Geiger, K. (2006). The effects of social comparison vers us mastery praise on children's intrinsic motivation. Motivation and Emotion 30 (4), 335 345. Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and Reading. Journal of Verbal Learning and Verbal Behavior, 19 450 466. Dunn ing, D., Heath, C., & Suls, J. M. (2004). Flawed self assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, 5(3), 69 106. Engle, R.W., Kane, M. J., & Tuholski, S.W. (1999). Individual differences in working memorycapacity and what they tell us aboutcontrolled attention, general fluid intelligence, and function of theprefrontal cortex. In A. Miyake & P. Shah (Eds.), Models of working memory: Mechanisms of active maintenance and executive contro l (pp. 102 134). New York: Cambridge University Press. Eysenck, M. W. (1992). Anxiety: The Cognitive Perspective. Hove: Lawrence Erlbaum Associates. Folkman, S., Lazarus, R. S., Dunkel Schetter, C., Delongis, A., & Gruen, R. (1986). Dynamics of a stressf ul encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality and Social Psychology, 50 992 1003. Gaab, J., Rohleder, N., Nater, U. M., & Ehlert, U. (2005). Psychological determinants of thecortisol stress response: The role of anticipatory cognitive appraisal. Psychoneuroendocrinology, 30, 599 610 Gimmig, D., Huguet, P., Caverni, J., & Cury, F. (2006). Choking under pressure and working memory capacity: When performance pressure reduces fluid intelligence. Psychonomic Bul letin & Review, 13, 1005 1010.

PAGE 41

41 Kelsey, R. M., Blascovich, J., Leitten, C. L., Schneider, T. R., Tomaka, J., & Wiens, S. (2000). Cardiovascular reactivity and adaptation to recurrent psychological stress: The moderating effects of evaluative observation. Ps ychophysiology, 37, 748 756. Klein, K., & Fiss, W. H. (1999). The reliability and stability of the Turner and Engle working memory task. Behavior Research Methods, Instruments, & Computers, 31 429 432. Lai., S. (2008). Competency threat: What are the e ffects on working memory? Lazarus, R. S., & Folkman, S. (1984). Stress, Appraisal and Coping. New York: Springer. Osborne, J. W. (2001). Testing stereotype threat: Does anxiety explain race and sex differences in achievement? Contemporary Educational Psychology, 26 291 310. Rhodewalt, F. (1990). Self handicappers: Individual differences in the preference for anticipatory self protective acts. In R. L. Higgins, C. R. Snyder, & S. Berg las (Eds.), Self (pp. 69 106). New York: Plenum. Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reducesworking memory capacity. Journal of Personality and Social Psychology, 85, 440 452 Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reduces working memory capacity. Journal of Personality & Social Psychology 85, 440 452. Schoofs, D., Preub, D., & Wolf, O. T. (2008). Psychosocial stress induces working me mory impairments in an n back paradigm. Psychoneuroendocrinology, 33, 643 653. Skinner, N., & Brewer, N. (2002). The dynamics of threat and challenge appraisals prior tostressful achievement events. Journal of Personality and Social Psychology, 83, 678 692. math performance Journal of Experimental Social Psychology, 35, 4 28. Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1970). Manual for the State Trait Anxi ety Inventory. Palo Alto, CA: Consulting Psychologists Press.

PAGE 42

42 Stangor, C., Carr, C., & Kiang, L. (1998). Activating stereotypes undermines task performance expectations. Journal of Personality and Social Psychology, 75 1191 1197. Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist, 52, 613 629. Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans. Journal of Personality and Social Psychology, 69, 797 811. Steele, C. M., Spencer, S., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. In M. Zanna (Ed.). Advances in experimental social psychology (Vol. 34). New Yor k: Academic Press. Thompson, T., & Dinnel, D. L. (2007). Is self worth protection best regarded as intentional self handicapping behavior or an outcome of choking under pressure? Educational Psychology, 27, 509 531. Tomaka, J., Blascovich, J., Kelsey, R. M., & Leitten, C. L. (1993). Subjective, physiological, and behavioral effects of threat and challenge appraisal. Journal of Personality and Social Psychology, 65, 248 260. Vick. S. B., Seery, M. D., Blascovich, J., & Weisbuch, M. (2008). The effect of gender stereotype activation on challenge and threat motivational states. Journal of Experimental Social Psychology, 44, 624 630.

PAGE 43

43 BIOGRAPHICAL SKETCH Ye Wang was born and grew up in Shanghai, China. She earned her Bachelor of Arts in Philosophy from Peking University in China in 2007. In fall 2007 she began her graduate study and majored in educational p sychology at the University of Florida in Gainesvi lle, Florida.