212 views

Uploaded by JORGE BARBOSA

- Math Anxiety Among University Students
- 2010 Learning Disability Guidelines Acc
- 23692397 Mind Masters
- Fact Sheet Assignment on Dyslexia
- Visual Storytelling Guide
- AP PSYC Essay Prompts & Scoring 2007
- Executive Function at Early Tese de Doutorado CLARK 2008
- Programme Handbook MScESLPLD Programme Handbook 2014-16 Cohort
- brain science reading
- educ 255 - multicultural paper - final 1
- Conduct and Behavior Problems
- HERMES_D61_CognitiveTrainingExercises_final.pdf
- huun
- eip process draft
- The Challenge When Navigating the Global is the Discovery of the Personal and the Universa1
- Dyslexia Handbook Teacherstrategies Kanth
- 6 ODEX.ppt
- Quiz
- working memory.pdf
- 1 How Concepts Are Acquired

You are on page 1of 9

C 2010 The Division for Learning Disabilities of the Council for Exceptional Children

Students with Learning Disabilities

Frances Prevatt and Theresa L. Welles

Florida State University

Huijun Li

Beth Israel Deaconess Medical Center and Harvard Medical School

Briley Proctor

Florida State University

The impact of memory and anxiety on math performance was analyzed in a sample of 115

college undergraduates, all of whom had a diagnosed learning disability. The direct effects of

memory and anxiety on math performance were first examined, followed by an examination

of whether anxiety moderates the relationship between memory and math. Both memory

and anxiety were found to directly affect math performance. Additionally, anxiety served as

a moderator of the relationship between memory and math for most, but not all, measures

of math achievement. The moderating effect of anxiety was stronger for long-term retrieval

than for short-term memory. The relationships between memory, anxiety, and math were not

significantly different for males and females. These findings suggest that, when working with

individuals who have low anxiety but poor memory, enhancing memory strategies may be

effective for remediating problems in math. However, for those with high levels of anxiety, it

may be more efficacious to first ameliorate the anxiety, as working on memory may have a

negligible effect on math performance for these individuals.

for students with disabilities (Individuals with Disabilities

Education Act, 1997, 2004). Concurrently, the number of The notion that the cognitive processing skills involving

students with learning disabilities who enter 2-year and 4- memory are significantly related to math performance in gen-

year colleges has increased dramatically (Henderson, 1999; eral is supported in the current literature (Beilock, Kulp, Holt,

Trainin & Swanson, 2005). Strawser and Miller (2001) re- & Carr, 2004; Casey, Nuttall, & Pezaris, 1997; Cassady, 2004;

ported that approximately 45 percent of students with mild Cirino, Morris, & Morris, 2007; Finney & Schraw, 2003;

learning disabilities who graduated from high school entered McGlaughlin et al., 2005; Miller & Bichsel, 2004; Trainin

a postsecondary institution. In fact, according to Trainin and & Swanson, 2005; Young, 2004). More specifically, studies

Swanson (2005), students with learning disabilities now rep- have found cognitive processing deficiencies in short-term

resent the largest group of college students with disabilities. and long-term memory for nondisabled college students who

However, students with learning disabilities may have a dif- performed poorly in math (Trainin & Swanson, 2005). Sev-

ficult time in college. Difficulty with college math appears eral studies on children experiencing math difficulties also

to be a likely reason for students with learning disabilities indicate that these individuals have difficulties storing and ac-

to drop out of college (McGlaughlin, Knoop, & Holliday, cessing math facts and knowledge from short-term and long-

2005). College students with learning disabilities spend a term memory (e.g., Bull, Johnston, & Roy, 1999; Hitch &

tremendous amount of time working on math; however, their McAuley, 1991; Swanson, 2006; Van de Sluis, Van de Leig,

severe deficits in math achievement persist, often leading & Jong, 2005). For example, Hitch and McAuley (1991)

to overall academic failure and attrition (Jones, Wilson, & explained such observations by suggesting that children with

Bhojwani, 1997). An understanding of the specific nature of math difficulties might have slower access to number rep-

math difficulties encountered by students with learning dis- resentations in long-term memory, which in turn may lead

abilities is important. It is likely that math deficits are linked to slow counting and low digit span that require short-term

to other factors, among them memory and anxiety. memory. Bull and Johnston (1997) also found that children

with low mathematics abilities performed significantly worse

than children with high ability in mathematics on short-term

Request for reprints should be sent to Frances Prevatt, PhD, Department

of Educational Psychology and Learning Systems, Stone 307 Florida State

memory tasks before reading ability was controlled. After

University, Tallahassee, Florida 32306. Electronic inquiries may be sent to reading ability was controlled, however, short-term memory

fprevatt@fsu.edu differences became insignificant between the two groups.

40 PREVATT ET AL.: MATH PERFORMANCE

What remained significant was the speed of identifying num- applied math performance for females, but not males. The

bers, which constitute representations that must be retrieved relationship between math ability and math anxiety for fe-

from long-term memory. One reason why processing speed males appears quite complex, and it is difficult to determine

differences are potentially important is that tests measur- the direction of the relationships, such as the relationship be-

ing processing speed are usually related to performance of tween math preparedness and math anxiety. For example, be-

short-term memory. It is unknown whether these types of cause math anxiety might reduce the number of math classes

findings for children are replicated in a college student pop- taken, the lack of preparation and experience in math might

ulation, when the complexity of the math tasks is at a higher increase the amount of anxiety in college students (Haynes,

level. Mullins, & Stein, 2004).

It is evident that there is empirical justification for the

study of memory, anxiety, and gender as they relate to the

Anxiety and Memory math achievement of college students, but little research has

been published on students with learning disabilities at the

Studies show that college students and adults with learning postsecondary level. This area is important, as enrollment

disabilities are highly affected by their level of anxiety in aca- in postsecondary institutions by students with documented

demic settings (Manglitz, Hoy, Gregg, King, & Moreland, learning difficulties is increasing. This study was designed to

1995). A number of studies have investigated the nega- evaluate the combined effect of memory and anxiety on the

tive impact of anxiety on memory capacity (Beilock et al., math performance of college students with diagnosed learn-

2004; Casey et al., 1997; Cassady, 2004; Kellogg, Hopko, ing disabilities. Although this work would be important for

& Ashcraft, 1999; Miller & Bichsel, 2004). For example, all students, it is especially important for those known to have

some research on adults and college students depicts the learning problems. Previous work has documented that there

negative impact of distracting thoughts and worries on math- are different effects for different types of math, for example,

ematics performance, resulting from disruption of the central Miller and Bichsel (2004) found differential effects for ap-

executive component of memory that is essential for com- plied and basic math. Based on previous research, it appears

plex problem solving (Beilock et al., 2004; Cassady, 2004; that utilizing only composite scores for math might cover im-

Miller & Bichsel, 2004). According to Beilock et al. (2004), portant fine-grained differences in the underlying processes

pressure-induced capacity limitations may result in perfor- that make up these constructs. Therefore, four different math

mance difficulties on tasks that require a sequence of mental subtests were utilized in this study, with the intention of

operations with demands on storage and processing (i.e., evaluating (a) fluency, (b) basic math facts, (c) knowledge

math calculation and problem solving). of symbols and concepts, and (d) the ability to analyze and

Anxiety has different manifestations and affects different solve math problems by reading and listening to a problem.

cognitive functions. For example, cognitive worry, as a ma- It was hypothesized that anxiety would moderate the rela-

jor component of anxiety, is manifested in negative self-talk tionship between memory and math. This effect was hypoth-

about one’s abilities. When college students are worried and esized to be stronger for long-term retrieval than short-term

anxious about studying or performing in academic situations, memory because long-term retrieval involves more compli-

they may be incapable of demonstrating their actual level of cated systems (e.g., associative memory, ideational fluency,

knowledge and skills (Weinstein & Palmer, 1990). Theo- meaningful memory, associative fluency, expressional flu-

ries of anxiety, including the processing efficiency theory ency, naming facility, and word fluency), and thus might be

(Eysenck & Calvo, 1992) and inhibition theory (Connelly, more susceptible to cognitive distractions associated with

Hasher, & Zacks, 1991) suggest that anxiety and worry about worry. Additionally, moderator effects were anticipated to be

performance will result in a reduction of memory capacity. more likely for the more complicated math tests (e.g., knowl-

Miller and Bichsel (2004) evaluated two types of memory in edge of symbols and concepts and the ability to analyze and

relationship to anxiety and math performance, verbal work- solve math problems by reading and listening to a problem).

ing memory and visual working memory; however, they did Solving basic math problems (e.g., math facts and fluency)

not evaluate long-term memory. were expected to utilize ingrained skills, less affected by

levels of anxiety.

Anxiety and memory were the primary variables being

Gender Differences in Math Performance

considered and therefore were analyzed together using a mod-

erator model. Due to the complexity of interpreting a three-

Some studies on gender differences in math performance

way interaction, gender effects were analyzed separately by

suggest that male college students tend to score higher than

evaluating whether the correlations among anxiety, memory,

females on tasks related to geometry and measurement and

and math differed for males and females. Due to inconsistent

females outperform male students in numbers and opera-

results in previous research, no directional hypotheses were

tions (Hagedorn, Siadat, Fogel, Nora, & Pascarella, 1999).

made regarding gender. Specific research questions were as

Furthermore, according to Miller and Bichsel (2004), gen-

follows:

der was found to moderate the relationship between anxiety

and math performance in college students, and this effect 1. Is there a gender effect for math achievement? Do the

differed depending on the type of math performance. For ex- correlations among anxiety, memory, and math differ

ample, math anxiety accounted for more of the variance in for males and females?

basic math skills for males than females, while math anx- 2. What are the direct relationships among different mea-

iety accounted for a significant amount of the variance in sures of memory (long term and short term), anxiety,

LEARNING DISABILITIES RESEARCH 41

and math (calculation skills, quantitative concepts, ap- follows: 48.7 percent Caucasian, 28.7 percent African Amer-

plied problems, and math fluency)? ican, 6.1 percent Hispanic, 2.6 percent Other, and 13.9 per-

3. Does anxiety serve as a moderator of the relationship cent Unknown. Twenty-five percent (n = 29) of the sample

between memory and math achievement? had a previous diagnosis of learning disabilities (learning

disability in writing, n = 3; learning disability in math, n =

4; learning disability in reading, n = 8; processing deficit,

METHOD n = 1; unknown area of disability, n = 13). Additionally,

15.7 percent (n = 18) reported a previous diagnosis of atten-

Participants tion deficit/hyperactivity disorder. The mean undergraduate

grade point average (GPA) of the sample was 2.40 on a

The sample consisted of 115 college students (56 percent 4-point scale (SD = .63). Other performance measures in-

male) who were diagnosed as having a learning disabil- cluded: SAT Verbal/Critical Reading score (M = 528, SD =

ity in math as a result of a comprehensive psychoeduca- 81); Quantitative/Math (M = 468, SD = 82); SAT Total (M

tional evaluation conducted at a clinic located on the cam- = 985, SD = 127), and GIA (M = 92, SD = 9.9).

pus of a large southeastern university. Trained master’s- and

doctoral-level students, under the supervision of licensed

psychologists, administered measures to assess cognitive, MEASURES

academic, and emotional functioning. In addition, the partic-

ipants completed informed consent documentation and pro- Math Achievement

vided demographic information. The clinic uses a modified

version of the aptitude-achievement consistency model to di- WJ-III ACH (Woodcock et al., 2001a) is an individually ad-

agnose learning disability, as developed by Flanagan, Keiser, ministered, norm-referenced test of academic achievement

Bernier, and Ortiz (2003). The aptitude-achievement consis- that may be administered to individuals from age 2 to over

tency model conceptualizes a learning disability as a specific 90. The instrument consists of 12 subtests within a standard

achievement deficit that can be explained by a corresponding battery, yielding composite standard scores in Broad Read-

deficit in one or more cognitive processes that are related to ing, Broad Mathematics, Broad Written Language, and Oral

the area of achievement deficit. In the university clinic, two Language clusters (Woodcock et al., 2001a). For the purpose

instruments were used to assess for learning disabilities: the of this study, the standard scores from the subtests for Cal-

Woodcock-Johnson III Tests of Cognitive Abilities (WJ-III culation, Applied Problems, Math Fluency, and Quantitative

COG; Woodcock, McGrew, & Mather, 2001b), and the Concepts were used. The subtests each have a mean of 100,

Woodcock-Johnson III Tests of Achievement (WJ-III ACH; and a standard deviation of 15.

Woodcock, McGrew, & Mather, 2001a). To meet the criteria

for a learning disability in math, a client had to first demon- Calculation

strate a weakness in one or more of the WJ III ACH math

clusters (e.g., Broad Math, Math Reasoning, Math Calcula- This test measures the ability to perform mathematical cal-

tion). Weakness was defined by a standard score below 85 culations that are presented in a traditional problem format.

(<16th percentile). In addition, the client had to demonstrate The median reliability is .85 in the 5–19 age range and .89

a weakness (<85) in one or more of the seven broad ability in the adult age range.

cluster scores from the WJ III COG (e.g., Comprehension-

Knowledge, Fluid Reasoning, Short-Term Memory, Long- Applied Problems

Term Retrieval, Processing Speed, Visual Processing, Au-

ditory Processing). The cognitive processing weakness also This test assesses an individual’s ability to analyze and solve

needed to be in an area that has been empirically linked to math problems by reading and listening to a problem, rec-

math performance (e.g., Fluid Reasoning, Comprehension- ognizing the required procedure and performing relatively

Knowledge). Finally, in order to ensure that the client was simple calculations. This test’s median reliability is .92 in the

not one with overall low cognitive ability, the client had to 5–19 age range and .95 in the adult range.

present with a General Intellectual Ability (GIA) score of at

least 85. The GIA is a composite score on the WJ III COG that

is analogous to a full scale IQ score; it is computed through Math Fluency

a differential weighting of the seven broad ability cluster

scores from the WJ III COG. Additional variables were also This test measures the ability to solve simple addition, sub-

considered during the diagnostic process, such as the client’s traction, and multiplication problems quickly within a pre-

history of math difficulties, the rule-out of other disorders scribed time limit. The median reliability of this test is .89 in

and conditions (e.g., anxiety, low motivation) that could be the 7–19 age range and .92 in the adult range.

the primary cause of the math difficulty, and the client’s re-

sponse to previous efforts to overcome the math difficulty

(e.g., attempt and response to tutoring, participating in study Quantitative Concepts

groups, meeting with the professor).

The mean age of the participants was 23.6 (SD = 8.9; This test consists of two subtests measuring the knowledge of

range 16–61). The race/ethnicity of the participants was as mathematical concepts, symbols, and vocabulary. This test

42 PREVATT ET AL.: MATH PERFORMANCE

has a median reliability of .90 in the 5–19 age range and .94 TABLE 1

in the adult range. Descriptive Statistics for the Study Measures

Females Males

Memory

Variable M (SD) Range M (SD) M(SD)

The WJ-III COG (Woodcock et al., 2001b) is an individually W-J III Test of Achievement

administered, norm-referenced test of cognitive functioning Applied 91.24 (7.89) 65–118 92.82 (8.41) 90.03 (7.28)

for individuals ranging in age from 2 to 90. The test con- Problems

sists of 10 subtests within a standard battery, yielding com- Calculation 91.67 (10.73) 63–119 92.64 (11.55) 90.92 (10.07)

posite standard scores in a variety of cognitive processing Math 87.12 (14.19 51–123 88.16 (11.80) 86.32 (15.83)

areas (clusters). For the purpose of this study, the compos- Fluency

ite standard scores in the areas of Short-Term Memory and Quantitative 90.74 (9.08) 68–114 92.14 (9.49) 89.73 (8.71)

Concepts

Long-Term Retrieval were used.

W-J III Tests of Cognitive Ability

Long-Term 92.42 (11.45) 45–139 92.02 (10.58) 92.72 (12.15)

Retrieval

Short-Term Memory

Short-Term 95.74 (14.21) 55–137 95.46 (15.93) 95.95 (12.85)

Memory

This cluster is an aspect of cognitive efficiency that measures LASSI 19.80 (7.89) 6–33 19.67 (7.72) 19.90 (8.07)

the ability to apprehend and hold information in immediate Anxiety

awareness and then use it within a few seconds. This cluster

Note: W-J III = Woodcock-Johnson III. LASSI = Learning and Study

has median reliabilities of .87 in the 5–19 age range and .92

Strategies Inventory.

in the adult range. The subtests that compose this cluster

include Numbers Reversed, a measure of working memory,

and Memory for Words, a measure of memory span.

shown to be positively correlated with GPA and is proposed

Long-Term Retrieval to be an effective tool for predicting academic performance

(Hulick & Higginson, 1989; Yip & Chung, 2002).

This cluster measures the ability to store information and

fluently retrieve it later in the process of thinking. This broad Data Analysis

ability requires the use of associative memory, ideational flu-

ency, meaningful memory, associative fluency, expressional Descriptive statistics for the study variables are given in

fluency, naming facility, and word fluency. The median relia- Table 1. As can be seen, the students in this sample had

bilities for this cluster are .87 in the 5–19 age range and .93 in mean math achievement scores on all four math tests that

the adult range. The two subtests composing the Long-Term were lower than the normative values. Using one-sample

Retrieval cluster include Visual-Auditory Learning, a mea- t tests, all four math tests were significantly lower than their

sure of associative memory, and Retrieval Fluency, a measure expected mean of 100 (p < .001 for all tests).

of ideational fluency.

Gender Effects

of gender on the four math outcomes. Overall, 56 percent

Learning and Study Strategies Inventory (LASSI; Weinstein of the sample were males. There were no significant main

& Palmer, 1990). The LASSI is a 10-scale assessment tool de- effects (p > .05 for all tests).

signed to measure students’ use of learning and study strate-

gies and methods. The subscale for anxiety addresses the Bivariate Relationships Among Memory, Anxiety,

degree to which students worry about school and their per- and Math Achievement

formance. The LASSI yields percentile scores; scores above

the 75th percentile indicate areas of success for the students; Table 2 shows the correlation matrix for the relationships

scores between the 50th and 75th percentile represent areas among the two measures of memory, the measure of anxiety,

that may need improvement. Scores under the 50th percentile and the four measures of math. Regarding memory, there

represent areas of significant weaknesses. In this study, raw were significant correlations between Long-Term Retrieval

scores were used. A higher raw score indicates lower levels and three of the four math subtests: Calculation, Applied

of anxiety. In our sample, the mean raw score was approx- Problems, and Quantitative Concepts. There were significant

imately 20, which corresponds to a percentile rank of 25; correlations between Short-Term Memory and all four of the

thus, overall, our sample experienced significant difficulties math subtests. Regarding anxiety, there were significant cor-

with anxiety. The user’s manual (Weinstein & Palmer, 1990) relations between LASSI anxiety and all four math subtests.

reports internal consistency (coefficient alpha) of .87 for the Finally, with regard to memory (two types) and anxiety, there

Anxiety scale. The norm group was composed of 1,092 col- was a significant correlation between Long-Term Retrieval

lege students from 12 different schools. The LASSI has been and anxiety.

LEARNING DISABILITIES RESEARCH 43

TABLE 2 TABLE 3

Correlations Among Memory, Anxiety, and Math Measures Regression Models for the Moderator Effects of Anxiety

1. AP Math Calculation

2. CL .66∗∗ Step 1 LTR .20 .02

3. MF .28∗∗ .43∗∗ Anxiety .19 .19 .00 .34 .00

4. QC .64∗∗ .62∗∗ .34∗∗ Step 2 LTR × Anxiety .26 .06 .00 .39 .00

5. LTR .44∗∗ .30∗∗ .10 .33∗∗ Math Fluency

6. STM .44∗∗ .26∗∗ .29∗∗ .40∗∗ .28∗∗ Step 1 LTR .02 .77

7. LA .29∗∗ .39∗∗ .28∗∗ .27∗∗ .26∗∗ .06 Anxiety .08 .08 .01 .27 .00

Step 2 LTR × Anxiety .08 .00 .79 .04 .74

Note: 1 = Applied Problems. 2 = Calculation. 3 = Math Fluency. 4 =

Applied Problems

Quantitative Concepts. 5 = Long-Term Retrieval. 6 = Short-Term Memory.

Step 1 LTR .38 .00

7 = LASSI Anxiety.

∗ p < .05, ∗∗ p < .01. Anxiety .23 .23 .00 .18 .03

Step 2 LTR × Anxiety .33 .10 .00 .48 .00

Quantitative Concepts

In order to further investigate the effects of gender, the

Step 1 LTR .27 .00

sample was subdivided and a series of correlations rerun for Anxiety .14 .14 .00 .20 .04

males and females. Specifically, correlations were run for Step 2 LTR × Anxiety .20 .06 .01 .37 .01

Long-Term Retrieval with all four math subtests, for Short- Math Calculation

Term Memory with all four math subtests, and for anxiety Step 1 Short-Term Memory .21 .01

with all four math subtests. A Fisher r to z transformation Anxiety .20 .20 .00 .38 .00

(two-tailed) was calculated to evaluate whether the correla- Step 2 STM × Anxiety .21 .01 .31 .16 .31

tion coefficients were significantly different across gender. Math Fluency

This analysis showed that none of the correlations were sig- Step 1 STM .27 .00

nificantly different for males compared to females (p > .05 Anxiety .15 .05 .00 .26 .00

for all tests). Step 2 STM × Anxiety .16 .01 .52 .11 .51

Applied Problems

Step 1 STM .43 .00

Anxiety as a Moderator

Anxiety .27 .27 .00 .26 .00

Step 2 STM × Anxiety .31 .03 .02 .34 .02

A moderator model would hypothesize that anxiety interacts Quantitative Concepts

with memory to produce an effect on math achievement. That Step 1 STM .37 .00

is, the relationship between memory and math performance Anxiety .21 .21 .00 .26 .00

would depend upon an individual’s level of anxiety. A signif- Step 2 STM × Anxiety .21 .00 .48 .12 .48

icant increase in the multiple R2 following the entry of the Note: LTR = Long-Term Retrieval. STM = Short-Term Memory.

interaction term into the regression analysis already contain-

ing the memory and anxiety variables would be indicative

of a moderator effect. Hierarchical regression was conducted anxiety. In the two other cases (Applied Problems and Quan-

in two sets; the first set utilized Long-Term Retrieval as the titative Concepts) there was a stronger effect for Long-Term

measure of memory, while the second set utilized Short-Term Memory.

Memory as the measure of memory. Each set ran a separate For the analyses with Short-Term Memory, anxiety served

analysis for each of the four math outcome measures. All as a moderator of the relationship between memory and math

variables were first centered around their means to facili- for one of the four math outcomes: Applied Problems. The

tate interpretation of the interaction terms and to control for model summaries can be seen in Table 3. The full model

multicollinearity (Aiken & West, 1991). Each regression fol- accounted for 31 percent of the variance in the Applied

lowed the same format: anxiety and memory were entered Problems math score. The interaction effect accounted for

in Step 1 and the interaction term (memory × anxiety) was 3 percent of the variance in math. Again, in most cases there

entered in Step 2. were significant direct effects for anxiety and memory, with

For the analyses with Long-Term Retrieval, anxiety served anxiety showing a greater effect for Math Calculation and

as a moderator of the relationship between memory and memory showing a greater effect for Applied Problems and

math for three of the four math outcomes: Calculation, Ap- Quantitative Concepts.

plied Problems, and Quantitative Concepts. The model sum- In order to understand the three significant interactions

maries can be seen in Table 3. The full models accounted for Long-Term Retrieval, these three regression models were

for a significant amount of the variance in math scores rerun separately based on level of anxiety, using noncen-

(Math Calculation–21 percent; Applied Problems–31 per- tered variables (Keith, 2006). The anxiety variable was di-

cent; Quantitative Concepts–21 percent). The interaction ef- chotomized into high and low, using a median split. The

fects accounted for 6, 10, and 6 percent of variance in math separate regression lines for high- and low-anxiety groups

scores, respectively. The unstandardized beta weights sug- illustrate the effect of the interaction. All three interactions

gested that, in general, there were significant direct effects followed the same pattern, only one visual representation will

for both memory and anxiety. In two cases (Math Calcu- be presented here, for illustrative purposes. Figure 1 displays

lation and Math Fluency) there was a stronger effect for the regression of Applied Math Problems on Long-Term

44 PREVATT ET AL.: MATH PERFORMANCE

on school-age children.

Because college students with learning disabilities are at

risk for both math difficulties and performance anxiety, it is

important to study these variables. Few studies have asked

whether anxiety may have a moderating effect on the relation-

ship between memory and math performance, using a variety

of outcome variables. Additionally, gender differences were

examined due to previous studies suggesting that males and

females perform differentially in different applications of

math.

Math achievement was represented by scores on the four

WJ-III math subtests of Calculation, Applied Problems, Math

Fluency, and Quantitative Concepts. Because the finding of

gender differences in math achievement would have affected

the subsequent analysis, we first examined whether males

FIGURE 1 Comparison of the regression slopes of applied math problems

on long-term memory for high- and low-anxiety groups.

and females had different means scores on any of the four

math subtests. The finding of no gender differences in this

study was surprising, given previous research in support of

Retrieval, for two levels of anxiety. With low anxiety, the gender differences (Hagedorn et al., 1999; Miller & Bichsel,

slope of the regression line is significant, indicating that 2004). All participants were clinic-referred college students

memory has a significant impact on math achievement. The who received a diagnosis of a math learning disability. There-

R2 for this model is .35 (p < .00). The unstandardized beta fore, we were not studying math performance in the general

weight (.50) for Long-Term Retrieval indicates that, for ev- population of students, but rather math performance of stu-

ery 1 point increase in memory, the average impact on math dents with diagnosed disabilities. It may be that the gender

achievement is a .5 point gain. Alternately, for the high- variance in math performance is lessened due to a restriction

anxiety group, the R2 is .05 (p > .05), indicating that mem- of range in this population.

ory does not predict math gains when anxiety is high. Similar Our initial analyses confirmed that this sample of college

patterns were seen for the other two math achievement ar- students did, in fact, have math deficits in all four areas. We

eas, Math Calculation and Quantitative Concepts. In both found evidence of direct effects of memory and anxiety on

cases, for the low-anxiety group one would predict close to math performance. Short-term memory was significantly cor-

a .5-point gain in math achievement with a 1-point gain in related with all four math subtests, while long-term memory

memory. Alternately, for the high-anxiety group, the slope was correlated with all math subtests except Math Fluency.

of the regression line is nonsignificant, indicating that Long- Also, academic anxiety (as measured by the LASSI anxiety

Term Retrieval did not predict math gains. scale) was correlated with all four math subtests. This find-

For Short-Term Memory, the same procedure was fol- ing extends the work of Trainin and Swanson (2005) who

lowed, although in this case the only significant interaction found direct effects for short-term and long-term memory on

was in the model predicting Applied Problems. First, the math in nondisabled students. In our sample, with only one

anxiety variable was dichotomized into high and low, using a exception (long-term memory and math fluency) we found

median split. Then, separate regression results for high- and significant effects for short-term memory on four types of

low-anxiety groups were examined to see the effect of the math, long-term memory on three types of math, and anxi-

interaction. Results indicated that, with low anxiety, memory ety on four types of math. Clearly, there are substantial and

has a significant impact on math achievement. The R2 for this pervasive effects across a broad array of specific areas of

model is .40 (p < .00). The unstandardized beta weight (.31) functioning. However, it is the interactions among these con-

for Short-Term Memory indicates that, for every 1-point in- structs that provides new information.

crease in memory, the average impact on math achievement Consistent with our hypotheses, several significant inter-

is a .3-point gain. For the high-anxiety group, the R2 is .17, actions were found. Results showed that anxiety moderated

(p < .00), indicating that memory still predicts math gains the effect of long-term retrieval on three of the four math

when anxiety is high, but accounts for less variance. The subtests: Math Calculation, Applied Problems, and Quan-

unstandardized beta weight (.20) for Short-Term Memory titative Concepts. In all three cases, there were significant

indicates that for every 1-point increase in memory, the av- direct effects, as well as significant moderator effects. After

erage impact on math achievement is a .2-point gain. dichotomizing the sample into high- and low-anxiety groups,

it was found that a moderately strong relationship continued

to exist between long-term retrieval and math performance

DISCUSSION when anxiety was low. However, when anxiety was high, the

relationship between memory and math performance was

The primary purpose of this study was to examine the effect nonsignificant. In the case of Math Fluency, there were no

of memory and anxiety on the math performance of col- moderator effects; in this instance, when anxiety and mem-

lege students diagnosed with a learning disability in math. ory were simultaneously evaluated, anxiety accounted for a

Although previous research has suggested relationships larger amount of variance than did memory. In other words,

between memory and math performance as well as anxiety for some types of math problems, both memory and anxiety

LEARNING DISABILITIES RESEARCH 45

are important, but low anxiety will protect one against the central executive system be freed up to attend to cognitive

effects of poor memory skills. Alternately, for Math Fluency, tasks.

if one has anxiety, then memory becomes less important; the It is interesting that anxiety as a moderator showed a

effect of the anxiety is so strong that it has a much more pow- stronger effect in conjunction with long-term retrieval than

erful influence than long-term memory. This makes sense, as short-term memory. For long-term retrieval, anxiety mod-

the Math Fluency test is a speeded test of relatively simple erated the relationship for three areas of math functioning:

math concepts, and anxiety is likely to have a strong impact math calculation, quantitative concepts, and applied prob-

when under time pressure. lems. Alternately, for short-term memory, anxiety served

For short-term memory, it was found that anxiety moder- as a moderator in only one area of math functioning: ap-

ated the effect of memory in only one case, Applied Prob- plied problems. We would speculate, based on the work of

lems. This effect was similar to the interactions found for Woodcock et al. (2001b), that long-term memory is associ-

long-term retrieval and indicated that a moderately strong ated with more complex cognitive processes, including asso-

relationship exists between short-term memory and Applied ciative memory, ideational fluency, meaningful memory, as-

Problems when anxiety was low. However, when anxiety was sociative fluency, expressional fluency, naming facility, and

high, the relationship between memory and Applied Prob- word fluency. These abilities are more likely utilized in the

lems was nonsignificant. There were direct effects of anxiety three math tests found to be moderated by long-term retrieval.

and memory for the remaining three math outcomes. For Math calculation requires solving problems in a variety of

Math Calculation, the effect for anxiety was stronger than formats, with increasing complexity. Applied problems re-

the effect for short-term memory. Alternately, for Quantita- quires the combination of several abilities, including oral

tive Concepts, the effect for short-term memory was stronger comprehension, decoding, recognition, and finally, problem

than the effect for anxiety. Finally, for Math Fluency, the two solving. Quantitative concepts also combine several abili-

effects (anxiety and short-term memory) were similar. ties, including use of symbols, vocabulary, and synthesis.

The results of this study support previous findings sug- We would speculate that these more complex skills will be

gesting a strong relationship between anxiety and math per- affected by the interplay of anxiety and long-term retrieval.

formance (Manglitz et al., 1995). Whereas previous studies Alternately, short-term memory involves somewhat simpler

have examined the way in which anxiety affects math perfor- abilities, more consistent with the skills required for the test

mance (e.g., anxiety affects memory, which in turn affects of math fluency. In this subtest, the problems are very sim-

math), this study took a different angle and asked whether ple addition, subtraction, and multiplication, with speed of

the well-established relationship between memory and math responding being the primary focus. Perhaps for the simple

is moderated by level of anxiety. Our findings are consistent tasks, level of anxiety does not overshadow memory skills.

with two complementary theories of anxiety: the processing The skills are so simple that anxiety does not have a detri-

efficiency theory (Eysenck & Calvo, 1992) and inhibition mental effect on one’s memory. Therefore, we see only a

theory (Connelly et al., 1991). Both theories suggest that direct effect for memory. Alternately, with the more complex

anxiety and worry about performance will result in a reduc- skills, variations in anxiety can cause enough disruption to

tion of memory capacity. The processing efficiency theory differentially impact memory.

identifies intrusive thoughts as creating a diversion of atten- In summary, understanding the relationships of anxiety

tional resources, thus reducing the available pool of working and memory to math achievement may help professionals

memory. With regard to math, we would speculate that the in- who work with or teach college students with learning dis-

dividual’s anxiety about their math performance serves as the abilities to target specific areas for intervention, as well as

diversionary stimulus. The inhibition theory alters this theory prevent failure of required math courses. Our study suggests

slightly by suggesting that the anxiety prohibits the individual that, when both anxiety and memory deficits are present in

from inhibiting attention to the anxious thoughts. The finding students with a math disability, anxiety reduction should gen-

of a positive moderator (i.e., interaction) effect has important erally be a first step, prior to interventions aimed at memory

implications for individuals struggling in math in spite of ad- strategies. For example, in the clinic where this study took

equate long-term memory. If a student comes in with low place, a specific anxiety reduction curriculum is offered as

anxiety, then working on memory strategies can have a direct an intervention for students. This 8-week behavioral program

impact on their math performance. In fact, for every one- combines academic consultation, relaxation, and systematic

point gain in memory, they might expect a half-point gain desensitization. This type of program has a history of posi-

in their math achievement. However, if they have concomi- tive results (e.g., Hembree, 1988; King & Ollendick, 1998);

tant anxiety, recommendations regarding memory strategies however, it might be beneficial to also screen students for

might not have as large an impact on math achievement. It memory skills and perhaps adding memory strategies as a

is recommended that anxiety reduction techniques be em- second phase of treatment. It will also be important to evalu-

ployed first, and only then will memory enhancement strate- ate the specific math area that is problematic for the student

gies likely result in concomitant math achievement gains. as the level of complexity of the math deficit may determine

The two theories described above help to explain why mem- the relative impact of anxiety versus memory deficits. From

ory enhancement strategies will be less effective in the face a research standpoint, future work with larger samples might

of anxiety. The intrusion of the diversionary stimulus (anx- do an even more fine-grained analysis, subdividing long-

iety about math) and the inability to inhibit the intrusion term retrieval and short-term memory into their component

keep the individual from utilizing memory strategies. Only subscales (e.g., Visual-Auditory Learning and Retrieval, Re-

by first decreasing the performance-related worries will the trieval Fluency, Numbers Reversed, and Memory for Words),

46 PREVATT ET AL.: MATH PERFORMANCE

in order to better understand the impact of different types of Henderson, C. (1999). College Freshman with disabilities. Washington, DC:

memory deficits. American Council on Education.

Hitch, G. J., & McAuley, E. (1991). Working memory in children with

specific arithmetic learning disabilities. British Journal of Psychology,

82, 375–386.

Hulick, C., & Higginson, B. (1989, November). The use of learning and study

REFERENCES strategies by college freshmen. Paper presented at the annual meeting

of the Mid-South Educational research Association, Little Rock, AR.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and inter- Jones, E. D., Wilson, R., & Hojwani, S. (1997). Mathematics instruction

preting interactions. Newbury Park, CA: Sage. for secondary students with learning disabilities. Journal of Learning

Beilock, S. L., Kulp, C. A., Holt, L. E., & Carr, T. H. (2004). More on Disabilities, 30, 151–163.

the fragility of performance: Choking under pressure in mathematical Keith, T. Z. (2006). Multiple regression and beyond. Boston: Pearson Edu-

problem solving. Journal of Experimental Psychology: General, 133, cation, Allyn and Bacon.

584–600. Kellogg, J. S., Hopko, D. R., & Ashcraft, M. H. (1999). The effects of time

Bull, R., & Johnston, R. S. (1997). Children’s arithmetical difficulties: Con- pressure on arithmetic performance. Journal of Anxiety Disorders, 13,

tributions from processing speed, item identification, and short-term 591–600.

memory. Journal of Experimental Child Psychology, 65, 1–24. King, N. J., & Ollendick, T. H. (1998) Empirically supported treatments for

Bull, R., Johnston, R. S., & Roy, J. A. (1999). Exploring the roles of the children with phobic and anxiety disorders. Journal of Clinical Child

visual–spatial sketch Pad and central executive in children’s arithmeti- Psychology, 27, 156–167.

cal skills: Views from cognition and developmental neuropsychology. Manglitz, E., Hoy, C., Gregg, N., King, M., & Moreland, C. (1995). The

Developmental Neuropsychology, 15, 421–442. relationship of depression and anxiety to ability and achievement in

Casey, M. B., Nuttall, R. L., & Pezaris, E. (1997). Mediators of gender adults with learning disabilities at a university and rehabilitation setting.

differences in mathematics college entrance test scores: A comparison Assessment in Rehabilitation and Exceptionality, 2, 163–178.

of spatial skills with internalized beliefs and anxieties. Developmental McGlaughlin, S. M., Knoop, A. J., & Holliday, G. A. (2005). Differentiating

Psychology, 33, 669–680. students with mathematics difficulty in college: Mathematics disabili-

Cassady, J. C. (2004). The impact of cognitive test anxiety on test com- ties vs. no diagnosis. Learning Disability Quarterly, 28, 223–232.

prehension and recall in the absence of external evaluative pressure. Miller, H., & Bichsel, J. (2004). Anxiety, working memory, gender, and math

Applied Cognitive Psychology, 18, 311–325. performance. Personality and Individual Differences, 37, 591–606.

Cassady, J. C. (2004). The influence of cognitive test anxiety across the Strawser, S., & Miller, S. P. (2001). Math failure and learning disabilities in

learning-testing cycle. Learning and Instruction, 14, 569–592. the postsecondary student population. Topics in Language Disorders,

Cirino, P. T., Morris, M. K., & Morris, M. K. (2007). Semantic, executive, 21, 68–84.

and visuospatial abilities in mathematical reasoning of referred college Swanson, H. L. (2006). Cognitive processes that underlie mathematical

students. Assessment, 14, 94–104. precociousness in young children. Journal of Experimental Child Psy-

Connelly, S. L., Hasher, L., & Zacks, R. T. (1991). Age and reading: The chology, 93, 239–264.

impact of distraction. Psychology and Aging, 6, 533–541. Trainin, G., & Swanson, H. L. (2005). Cognition, metacognition, and

Eysenck, M. W., & Calvo, M. G. (1992). Anxiety and performance: achievement of college students with learning disabilities. Learning

The processing efficiency theory. Cognition and Emotion, 6, 409– Disability Quarterly, 28, 261–272.

434. Van de Sluis, S., Van de Leij, A., & de Jong, P. F. (2005). Working memory

Finney, S. J., & Schraw, G. (2003). Self-efficacy beliefs in college in Dutch children with reading- and arithmetic-related LD. Journal of

statistics courses. Contemporary Educational Psychology, 28, 161– Learning Disabilities, 38, 207–221.

186. Weinstein, C. E., & Palmer, D. R. (1990). LASSI-HS user’s manual. Clear-

Flanagan, D. W., Keiser, S., Bernier, J., & Ortiz, S. O. (2003). Diagnosis of water, FL: H & H Publishing Company.

learning disabilities in adulthood. Boston: Allyn and Bacon. Woodcock, R. W., McGrew, K. S., & Mather, N. (2001a). Woodcock-Johnson

Hagedorn, L. S., Siadat, M. V., Fogel, S. F., Nora, A., & Pascarella, E. T. III tests of achievement. Itasca, IL: Riverside Publishing.

(1999). Success in college mathematics: Comparisons between reme- Woodcock, R. W., McGrew, K. S., & Mather, N. (2001b). Woodcock-Johnson

dial and nonremedial first-year college students. Research in Higher III tests of cognitive ability. Itasca, IL: Riverside Publishing.

Education, 40, 261–284. Yip, M. C. W., & Chung, O. L. L. (2002). Relation of study strategies to the

Haynes, A. F., Mullins, A. G., & Stein, B. S. (2004). Differential mod- academic performance of Hong Kong University students. Psycholog-

els for math anxiety I male and female college students. Sociological ical Reports, 90, 338–340.

Spectrum, 24, 295–318. Young, C. J. (2004). Contributions of metaknowledge to retrieval of natural

Hembree, R. (1988). Correlated causes, effects and treatments of test anxi- categories in semantic memory. Journal of Experimental Psychology:

ety. Review of Educational Research, 58, 47–77. Learning, Memory and Cognition, 30, 909–916.

Frances Prevatt is a professor in the Department of Educational Psychology and Learning Systems at Florida State University.

She is also codirector of the Adult Learning Evaluation Center. She earned her degree in clinical psychology from the University

of Virginia. Her research interests include attention deficit/hyperactivity disorder (ADHD) and learning disabilities in college

students, as well as motivational aspects of academic success in college students.

Theresa Welles is currently a licensed school psychologist in the state of Florida and a doctoral candidate in the Combined

Counseling and School Psychology Program at Florida State University. Her research interests are in the areas of motivation,

anxiety, ADHD, and learning disabilities in a college population of students.

LEARNING DISABILITIES RESEARCH 47

Huijun Li is Director of Multicultural Research at Commonwealth Research Center, Department of Psychiatry, Beth Israel

Deaconess Medical center, harvard Medical School. She earned her degree in school psychology from the University of Arizona.

Her research interests include emotional well-being of children and adolescents with disabilities and from different cultural

backgrounds, as well as cultural relevant interpretations of psychiatric illnesses.

Briley Proctor is an associate professor in the Department of Educational Psychology and Learning Systems at Florida State

University. She is also codirector of the Adult Learning Evaluation Center. She earned her degree in school psychology from the

University of Florida. Her research interests include ADHD and learning disabilities in college students, as well as educational

policy and advocacy.

- Math Anxiety Among University StudentsUploaded byDanny Davis
- 2010 Learning Disability Guidelines AccUploaded bypauladest
- 23692397 Mind MastersUploaded byReteshkumarverma
- Fact Sheet Assignment on DyslexiaUploaded byMBCass
- Visual Storytelling GuideUploaded byVisualStoryteller
- AP PSYC Essay Prompts & Scoring 2007Uploaded byKerilyn Wuerz
- Executive Function at Early Tese de Doutorado CLARK 2008Uploaded byCleumanicardi
- Programme Handbook MScESLPLD Programme Handbook 2014-16 CohortUploaded byKerryKeys
- brain science readingUploaded byapi-292667141
- educ 255 - multicultural paper - final 1Uploaded byapi-324504809
- Conduct and Behavior ProblemsUploaded bynao
- HERMES_D61_CognitiveTrainingExercises_final.pdfUploaded byVazia Rahma Handika
- huunUploaded bySunn Flower
- eip process draftUploaded byapi-340929370
- The Challenge When Navigating the Global is the Discovery of the Personal and the Universa1Uploaded byLeana Litchfield
- Dyslexia Handbook Teacherstrategies KanthUploaded byvarsham2
- 6 ODEX.pptUploaded byJd Jamolod Pelovello
- QuizUploaded byfreeloader999
- working memory.pdfUploaded byJudithBendezú
- 1 How Concepts Are AcquiredUploaded byGowri Nagamah
- brenna mcconnell june 10th iep assignmentUploaded byapi-243019491
- Head on June 2019 HandoutUploaded byStacy Ann Vergara
- learning disablity article review-finalUploaded byapi-399107713
- 65_PDFUploaded byMary Jhane Villanueva
- pat 2-electronic resource collectionUploaded byapi-313808473
- 94text3Uploaded byxinying94
- second draftUploaded byapi-238741224
- Ofsted - Full Inspection Report June 2008Uploaded bybrog
- InglêsUploaded bySuelen Daniel
- 21.-Conditional-Worksheet.docUploaded byKatia Leliakh

- Sensação e PercepçãoUploaded byJORGE BARBOSA
- Pensamento e RaciocínioUploaded byJORGE BARBOSA
- MemóriaUploaded byJORGE BARBOSA
- Linguagem e LeituraUploaded byJORGE BARBOSA
- Psicologia CognitivaUploaded byJORGE BARBOSA
- PHILOSOPHY IN FRAGMENTS: CULTIVATING PHILOSOPHIC THINKING WITH THE PRESOCRATICS - DANIEL SILVERMINTZUploaded byJORGE BARBOSA
- Psicologia Cognitiva 2Uploaded byJORGE BARBOSA
- Tomadas de Decisão de Risco em Adolescentes - Perspectiva EcológicaUploaded byJORGE BARBOSA
- Conference BrochureUploaded byJORGE BARBOSA
- Artigo sobre MemóriaUploaded byJORGE BARBOSA
- CRPD Guidance Document-EnglishUploaded byJORGE BARBOSA
- Semantics vs. world knowledge in prefrontal cortexUploaded byJORGE BARBOSA

- Ramesh and GargiUploaded byjayavignesh
- Goal SettingUploaded bySoetan Olawale Rasheed
- U Burke Litwin Model Chapters 9 and 10Uploaded byRamona Moraru
- HR CompetencyUploaded byStacey Brooks
- A Guide to the Fixed Mindset v the GrowthUploaded byholea
- Holiday Decision Making FinalUploaded bysonakshi29
- Police StressUploaded byIqbal Baryar
- ARM Lecture Eight(3)Uploaded bySaadiya Razzaq
- ConstructivismUploaded bydeepak manthan
- resumUploaded byapi-249776409
- Educ 2220 Lesson PlanUploaded byjthompson102
- task sheet e portfolioUploaded byapi-238909471
- EFFICACY OF LEADERSHIP ETHICS AS A TOOL FOR DEVELOPING CAPABILITIES: A STUDY ON MBA STUDENTSPERSPECTIVESUploaded byMohammad Miyan
- Transformational.pdfUploaded byOasis International Consulting Ltd
- 21st century learning design innovationUploaded byapi-240435029
- Community OrganizingUploaded byRyan Ignacio
- Salesperson Behavour Questionnair_replyUploaded byJohnarcher1983
- Words of GratitudeUploaded byCharlad Cha Bangtan Potter
- Organizational Citizenship Behaviour: The Role of Perceived Organizational Justice and Self-esteem.Uploaded byIOSRjournal
- Guidelines - Inferential Questions AssessmentUploaded byDavid Woo
- SINGER Practical EthicsUploaded byAstrid Busekist Sadoun
- Interpersonal CommunicationUploaded byeginabila
- OBUploaded byNishantraje Turekar
- Adaptation ModelUploaded byAnn Michelle Tarrobago
- Behavioral Finance IntroductionUploaded byNikhilSharma
- Social Studies Thesis ReferenceUploaded byJames Pino
- Chapter 4musUploaded bysrojas5
- Rubric Lesson RefreshUploaded bymounlakay
- assessment item 1 emm410Uploaded byapi-324846334
- culminating reflectionUploaded byapi-252901681