ArticlePDF Available

Development and Psychometric Evaluation of a Measure of Intuitive Eating

April 2006
Journal of Counseling Psychology 53(2):226-240

DOI:10.1037/0022-0167.53.2.226

Authors:

The Ohio State University

Intuitive eating is characterized by eating based on physiological hunger and satiety cues rather than situational and emotional cues and is associated with psychological well-being. This study reports on the development and initial psychometric evaluation of the Intuitive Eating Scale (IES) with data collected in 4 studies from 1,260 college women. Exploratory factor analysis uncovered 3 factors: unconditional permission to eat, eating for physical rather than emotional reasons, and reliance on internal hunger/satiety cues; confirmatory factor analysis suggested that this 3-factor model adequately fit the data after 4 items with factor loadings below .45 were deleted. IES scores were internally consistent and stable over a 3-week period. Supporting its construct validity, IES scores were (a) negatively related to eating disorder symptomatology, body dissatisfaction, poor interoceptive awareness, pressure for thinness, internalization of the thin ideal, and body mass; (b) positively related to several indexes of well-being; and (c) unrelated to impression management. (PsycINFO Database Record (c) 2012 APA, all rights reserved)

Means, Standard Deviations, and Correlations Among the Measures of Study 2 (N 476)

…

No caption available

…

Figures - uploaded by Tracy L. Tylka

Content may be subject to copyright.

Content uploaded by Tracy L. Tylka

Content may be subject to copyright.

Development and Psychometric Evaluation of a Measure

of Intuitive Eating

Tracy L. Tylka

Ohio State University

Intuitive eating is characterized by eating based on physiological hunger and satiety cues rather than

situational and emotional cues and is associated with psychological well-being. This study reports on the

development and initial psychometric evaluation of the Intuitive Eating Scale (IES) with data collected

in 4 studies from 1,260 college women. Exploratory factor analysis uncovered 3 factors: unconditional

permission to eat, eating for physical rather than emotional reasons, and reliance on internal hunger/

satiety cues; confirmatory factor analysis suggested that this 3-factor model adequately fit the data after

4 items with factor loadings below .45 were deleted. IES scores were internally consistent and stable over

a 3-week period. Supporting its construct validity, IES scores were (a) negatively related to eating

disorder symptomatology, body dissatisfaction, poor interoceptive awareness, pressure for thinness,

internalization of the thin ideal, and body mass; (b) positively related to several indexes of well-being;

and (c) unrelated to impression management.

Keywords: intuitive eating, eating behaviors, college women, assessment, psychometrics

In the field of psychology, the study of eating behaviors largely

has been a pathology-focused endeavor because it has explored

and identified correlates and predictors of disordered rather than

adaptive eating. It has been argued that psychologists also need to

consider adaptive behaviors that contribute to and maintain overall

psychological health (Seligman & Csikszentmihalyi, 2000). Some

psychologists have addressed adaptive eating behaviors when ex-

ploring unrestrained eating (e.g., Kahan, Polivy, & Herman, 2003;

Polivy & Herman, 1999) and the eating disorder continuum (e.g.,

Mintz & Betz, 1988; Tylka & Subich, 1999, 2004); however, they

have defined these behaviors as the absence of eating disorder

symptoms. Professionals within other disciplines (e.g., dietetics)

have discussed adaptive eating, but it usually is presented in the

form of behavioral guidelines for the intake of certain food groups

rather than an exploration of its correlates and predictors, and it is

often kept separate from the literature on clinical eating disorders

(Ogden, 2003). As a result, the study of eating behaviors is

disjointed, and much remains unknown about positive eating

behaviors.

Instruments assessing eating behaviors have mirrored this focus

on pathology. It is true that low levels and asymptomatic rank on

such measures as the Eating Attitudes Test–26 (EAT-26; Garner,

Olmsted, Bohr, & Garfinkel, 1982), the Eating Disorder

Inventory–3 (Garner, 2004), and the Questionnaire for Eating

Disorder Diagnoses (Mintz, O’Halloran, Mulholland, & Schnei-

der, 1997) are indicators of adaptive eating because they reflect the

absence of characteristics associated with clinical eating disorders

(e.g., preoccupation with food, binge eating, dietary restriction).

Yet, some scholars (e.g., Tribole & Resch, 1995) have argued that

adaptive eating is more than just the absence of these characteris-

tics because different internal cues are used to determine when,

what, and how much to eat. Individuals who eat adaptively often

use physiological hunger and satiety cues to guide their eating

behaviors, whereas individuals with clinical eating disorders often

use emotional cues to guide their eating behaviors. Also, individ-

uals may not display eating disorder symptomatology per se but

may not eat adaptively (e.g., by habitually eating in the absence of

hunger but not an amount large enough to be considered a binge or

by eating everything on the plate with no regard to satiety level).

Thus, adaptive eating may be negatively related to but not solely

defined by the absence of eating disorder symptoms. In order to

understand this construct more comprehensively, measures of

adaptive eating need to be developed. Use of these measures can

raise awareness about potential variables that can serve as buffers

to developing disordered eating behaviors. The purpose of the

present study, then, was to develop and assess the psychometric

properties of such a measure.

Although there could be several facets of adaptive eating, the

current study focuses on describing and measuring intuitive eating

(i.e., eating based on physiological hunger and satiety cues rather

than external and emotional cues; Tribole & Resch, 1995). Several

psychologists (e.g., Carper, Fisher, & Birch, 2000; Fedoroff, Po-

livy, & Herman, 1997; Polivy & Herman, 1987, 1992) and nutri-

tionists (e.g., Tribole & Resch, 1995) have argued that this style of

eating is adaptive because it is associated with a strong connection

with, understanding of, and response to internal physiological

needs pertaining to hunger and satiety as well as low preoccupa-

tion with food. These scholars have identified three central features

of intuitive eating: (a) unconditional permission to eat when hun-

gry and what food is desired, (b) eating for physical rather than

emotional reasons, and (c) reliance on internal hunger and satiety

cues to determine when and how much to eat. These components

Correspondence concerning this article should be addressed to Tracy L.

Tylka, Department of Psychology, Ohio State University, 1465 Mt. Vernon

Avenue, Marion, OH 43302. E-mail: tylka.2@osu.edu

2006, Vol. 53, No. 2, 226–240 0022-0167/06/$12.00 DOI: 10.1037/0022-0167.53.2.226

226

are interrelated, and the presence of each is necessary to reflect

intuitive eating (Tribole & Resch, 1995). A detailed discussion of

each feature, including why it is considered adaptive, is presented

next.

Unconditional Permission to Eat (When Hungry and What

Food Is Desired)

Unconditional permission to eat reflects a readiness to eat in

response to internal physiological hunger signals and the food that

is desired at the moment (Tribole & Resch, 1995). Individuals who

engage in this eating strategy do not try to ignore their hunger

signals, nor do they classify food into acceptable and nonaccept-

able categories and attempt to avoid food in the latter category.

Laboratory experiments have revealed that these individuals eat

more food after a period of not eating or after eating a small

amount of food rather than after eating a large amount of food

(Herman & Polivy, 1988; Woody, Costanzo, Leifer, & Conger,

1981). Therefore, their eating is controlled by their hunger and

satiety signals, unlike those who allow themselves to eat but have

no control over their eating (e.g., individuals who engage in binge

eating).

People who place conditions on when, how much, and what

foods they can eat (i.e., by restricting the timing, amount, and type

of food eaten according to some external standard) increase their

likelihood of feeling deprived and preoccupied with food (Polivy

& Herman, 1999; Tribole & Resch, 1995). When men who had no

prior preoccupation with food were placed on a greatly reduced

calorie diet for 6 months, many became extremely preoccupied

with food and engaged in binge eating that persisted even after the

diet was terminated (Keys, Brozek, Henschel, Mickelsen, & Tay-

lor, 1950). Children whose parents restricted their food intake were

more likely to eat in the absence of hunger and had higher body

mass than children whose parents did not substantially restrict their

food intake (Faith, Scanlon, Birch, Francis, & Sherry, 2004).

Laboratory experiments have revealed that restrained eaters over-

indulge in food as a result of perceiving that dietary rules have

been broken or that they have eaten a forbidden food (Herman &

Polivy, 1988; Woody et al., 1981). In addition, people engaging in

dietary restraint are more likely to allow visual and olfactory cues

of foods to guide their food intake than are people who do not

restrict their eating (Fedoroff et al., 1997). Because dietary re-

straint further increases food preoccupation (Keys et al., 1950;

Polivy & Herman, 1999), people who restrict their food intake

actually may eat more than people who give themselves uncondi-

tional permission to eat.

Eating for Physical Rather Than Emotional Reasons

People who eat intuitively use food to satisfy a physical hunger

drive and not to cope with their emotional fluctuations and/or

distress (Tribole & Resch, 1995). A boundary model has been

proposed that explores the connection between eating behavior and

emotions (Herman & Polivy, 1983). Individuals who do not diet

have two boundaries corresponding to hunger and satiety. When

hungry, they will eat so as to escape the hunger zone and will stop

eating when indifferent or slightly sated. Research has shown that

individuals who do not diet often eat less when they are anxious or

stressed than when they are calm, perhaps because of the appetite-

suppressing sympathomimetic effects of these emotions (Herman,

Polivy, Lank, & Heatherton, 1987). However, the behavior of

individuals who restrict their eating is largely under the control of

a third and unnatural diet boundary (Herman & Polivy, 1983).

When the diet boundary is breached, eating often becomes disin-

hibited and in defiance of the hunger and satiety boundaries.

Emotional agitation frequently disrupts the diet boundary because

people who restrain their eating increase their food intake when

they experience negative affect (Costanzo, Reichmann, Friedman,

& Musante, 2001; Herman et al., 1987).

Reliance on Internal Hunger and Satiety Cues to

Determine When and How Much to Eat

People who engage in intuitive eating are both aware of their

internal hunger and satiety signals and trust these signals to guide

their eating behavior (Carper et al., 2000; Tribole & Resch, 1995).

Awareness of inner experiences is a central aspect to well-being

(Rogers, 1964). This awareness is inborn; with development, how-

ever, some people substitute external rules (e.g., proscribing when,

what, and how much to eat) for inner experience as they internalize

societal messages that dieting will lead to favorable outcomes.

Laboratory experiments have revealed that young children have an

internal mechanism that helps them fairly accurately regulate food

intake; even though their intake at each meal was highly variable,

their total daily energy intake was relatively constant (e.g., a

high-calorie meal was followed by a low-calorie meal; Birch,

Johnson, Andresen, Petersen, & Schulte, 1991).

Many caregivers observe children’s variable eating behaviors,

conclude that children cannot adequately regulate food intake, and

adopt coercive strategies to exert control over children’s eating

behaviors (Birch et al., 1991). These strategies are counterproduc-

tive because they replace innate internal hunger and satiety signals

with external rules and lead to a disconnection from internal

experience and innate ability to regulate food intake. In fact, this

type of pressure in child feeding is associated with the emergence

of dietary restraint, weight gain, eating in the absence of hunger,

and eating in response to emotions (e.g., sadness or boredom) and

situational factors (e.g., the mere presence of food) among young

girls (Birch & Fisher, 2000; Birch, Fisher, & Davison, 2003;

Carper et al., 2000).

To date, an instrument has not been developed that assesses

these key components of intuitive eating. In four studies, the

current investigation reports the development of such a measure,

the Intuitive Eating Scale (IES), and its preliminary psychometric

examination with college women.

Study 1: Exploratory Factor Analysis and Construct

Validity

In Study 1, the IES items were developed. Once a scale is

developed, it is imperative to explore its factor structure and

determine whether it yields reliable and valid scores (Walsh &

Betz, 2001). Also, it is essential to determine whether the scale’s

items are internally consistent and whether the scale is strongly

related to other scales that measure similar and related constructs

in order to establish its integrity (Walsh & Betz, 2001). Several

hypotheses were generated. First, a three-factor structure corre-

227

INTUITIVE EATING SCALE

sponding to the central aspects of intuitive eating was expected to

emerge in the exploratory factor analysis. Second, scores on the

total IES and the emergent factor subscales were expected to

demonstrate evidence of internal consistency reliability. Third,

because intuitive eating is conceptualized as adaptive eating (Tri-

bole & Resch, 1995) and eating disorder symptomatology is con-

ceptualized as maladaptive eating (Garner, 1991), the IES was

expected to be related in a negative direction to eating disorder

symptomatology. Fourth, it has been asserted (e.g., Tribole &

Resch, 1995) that significant others can foster intuitive eating in

women by accepting their body size and by not pressuring them to

become thinner; as a result, women are more likely to refrain from

internalizing the thin-ideal prototype advocated by society, accept

their bodies, be aware of and attend to their internal hunger and

satiety signals, and eat in response to these signals. Indeed, it has

been found that women who do not restrict their food intake report

lower pressure for thinness, body dissatisfaction, and internaliza-

tion of the thin-ideal image than do women who restrict their food

intake (Mills, Polivy, Herman, & Tiggemann, 2002). Given the

above theory and research, the IES was hypothesized to be nega-

tively related to body dissatisfaction, poor interoceptive aware-

ness, pressure for thinness, and internalization of the thin-ideal

stereotype.

Method

Participants and procedure. Women enrolled in introductory psychol-

ogy courses volunteered to participate though the psychology department’s

organized research program. The study was described as an investigation of

the relation between eating habits, body attitudes, and personality charac-

teristics. After women were guaranteed anonymity and their informed

consent was obtained, they completed the questionnaires in a classroom

used as a research lab. The measures were counterbalanced to control for

order effects. Participants received credit that was applied toward their

class grade.

Responses from 5 women who did not complete at least 90% of any

given measure were not entered into the data set. The final data set included

responses from 391 women from a large Midwestern university who

ranged in age from 17 to 61 years (M⫽20.85, SD ⫽6.21). Women

identified as Caucasian American (87.7%), Asian American (3.8%), mul-

tiracial (3.4%), African American (3.1%), Native American (2.8%), and

Latina (0.5%). One woman (0.3%) did not indicate her ethnicity. Over half

of the women were freshmen (64.7%); of the remaining women, 15.3%

were sophomores, 8.2% were juniors, 9.0% were seniors, 0.8% were

postbaccalaureate students, 0.5% were graduate students, and 1.3% did not

specify their college rank. Participants described themselves as middle

class (50.6%), upper middle class (30.7%), working class (15.3%), and

upper class (2.3%). One percent did not report a socioeconomic

identification.

Measures. Twenty-eight items were designed to assess the following

key aspects of intuitive eating: (a) unconditional permission to eat when

hungry and what food is desired at the moment (12 items), (b) eating for

physical rather than emotional reasons (9 items), and (c) reliance on

internal hunger and satiety cues to determine when and how much to eat (7

items). Items were written until it was determined that the group of items

comprehensively and adequately reflected the central characteristics of

each of the three aspects of intuitive eating. To formulate the items, I used

information gathered from extant theoretical and empirical literature dis-

cussing intuitive eating, unrestrained eating, and how to promote adaptive

eating (e.g., Carper et al., 2000; Mills et al., 2002; Polivy & Herman, 1992;

Tribole & Resch, 1995). The response format for the IES is a 5-point

Likert-type scale (i.e., 1 ⫽strongly disagree,2⫽disagree,3⫽neutral,

4⫽agree,5⫽strongly agree). Higher scores indicate higher levels of

intuitive eating.

Following initial item generation, several steps were taken to ensure the

integrity of each item. One doctoral student in an American Psychological

Association-accredited counseling psychology program and two under-

graduate senior honors students provided feedback regarding the content,

clarity, and parsimony of each item. A counseling psychologist and a

nutrition consultant also were consulted to ensure that the items accurately

reflected the content domain. A majority of these evaluators agreed that all

items should be retained; however, three items were reworded for clarifi-

cation. No new items were proposed.

In addition to the IES, the following measures were given to the

participants. These scales typically have been used in research to assess

their respective constructs (Tylka & Subich, 2004).

The EAT-26 (Garner et al., 1982) was used to assess eating disorder

symptomatology. Each of its 26 items (e.g., “I am terrified about being

overweight”) is rated on a scale ranging from 1 (never)to6(always).

Although Garner et al. recommended that the responses never, rarely, and

sometimes receive a score of 0 and that the responses often, very often, and

always receive scores of 1, 2, and 3, respectively, participants’ total scores

were equal to the average of the coded responses to prevent range restric-

tion. Other researchers (e.g., Mazzeo, 1999; Tylka & Subich, 2004) also

have used this continuous scoring method with nonclinical samples of

college women. Items were averaged to obtain a total score; higher scores

indicate greater eating disorder symptomatology. Among college women,

a coefficient alpha of .91 and a 3-week test–retest reliability estimate of .80

for its scores have been reported (Mazzeo, 1999). For the present study,

alpha was .91 for its scores. The EAT-26 has been shown to be related to

the Drive for Thinness subscale (r⫽.84) and to the Bulimia subscale (r⫽

.55) of the Eating Disorder Inventory–2 (EDI-2; Garner, 1991) among

college women, supporting its convergent validity (Brookings & Wilson,

1994).

The Body Dissatisfaction (BD) and Interoceptive Awareness (IA) sub-

scales of the EDI-2 (Garner, 1991) were used to measure their respective

constructs. For the purposes of this study, only these two subscales of the

EDI-2 were administered. The BD subscale contains 9 items that measure

dissatisfaction with overall body size and the belief that parts of the body

are too large, whereas the IA subscale contains 10 items that assess poor

awareness of internal body states, such as emotions, hunger, and satiety.

Items are rated on a scale ranging from 1 (never true of me)to6(always

true of me). Although Garner (1991) recommended that the item responses

never true of me, seldom true of me, and sometimes true of me receive a

score of 0 and the responses often true of me, very often true of me, and

always true of me receive scores of 1, 2, and 3, respectively, the coded

responses were averaged to prevent range restriction. Other researchers

(e.g., Tylka & Subich, 2004) also have used this scoring method with

college women. Higher subscale scores are indicative of greater body

dissatisfaction and poorer interoceptive awareness, respectively. Among

college women, alphas of .91 for BD subscale scores and .78 for IA

subscale scores have been reported (Brookings & Wilson, 1994). For the

present study, alphas were .93 for BD and .86 for IA subscale scores. Over

a 3-week period, test–retest reliability estimates have been found to be .97

for BD and .85 for IA subscale scores among college women (Wear &

Pratz, 1987). Also, BD scores were related to body preoccupation (r⫽

.83), and IA scores were related to difficulty identifying feelings (r⫽.78)

among college women, supporting the convergent validity for these two

subscales (Tylka & Subich, 2004).

The Perceived Sociocultural Pressures Scale (Stice, Ziemba, Margolis,

& Flick, 1996) was used to determine the women’s reported pressure for

thinness from significant others (i.e., family, friends, and partners) and the

media. It contains eight items (e.g., “I’ve felt pressure from my family to

lose weight”) that are each rated along a scale ranging from 1 (strongly

228 TYLKA

disagree)to5(strongly agree) and are averaged to obtain a total score.

Higher scores reflect greater perceived pressure to be thin. Among high

school and college women, alpha for the Perceived Sociocultural Pressures

Scale scores was .87, its stability of its scores over a 2-week period was

.93, and it was related (r⫽.51) to retrospective reports of pressure to lose

weight during childhood (Stice et al., 1996). For the present study, alpha

was .88 for its scores.

The Internalization subscale of the Sociocultural Attitudes Toward Ap-

pearance Questionnaire (Heinberg, Thompson, & Stormer, 1995) was used

to assess the extent to which the women believed that the thin-ideal societal

stereotype is the ideal body type. It contains eight items (e.g., “Photographs

of thin women make me wish that I were thin”) that are each rated along

a scale ranging from 1 (completely disagree)to5(completely agree) and

are averaged to obtain a total subscale score. Higher subscale scores

indicate greater internalization of the thin-ideal stereotype. Among college

women, alpha was .88 for its scores, and its unidimensionality was upheld

(Heinberg et al., 1995). This subscale was related (r⫽.64) to college

women’s perceptions of ideal body type (Tylka & Subich, 2004). For the

present study, alpha was .90 for its scores.

Results and Discussion

Data first were examined to ensure that the variables’ distribu-

tions would not violate statistical assumptions of the planned

analyses; no substantial violations were uncovered. Further, the

sample size exceeded the number of cases needed for a

participants-to-parameter ratio of 5–10:1, which is required to

accurately estimate the factor structure of a scale (Bentler, 1990).

Means, standard deviations, and intercorrelations of the measures

are included in Table 1.

Exploratory factor analysis. SPSS Version 13.0 was used to

conduct the exploratory factor analysis and all other statistical

analyses reported in this article, except for the confirmatory factor

analysis reported in Study 2. The significance of Bartlett’s test of

sphericity,

␹

(300, N⫽391) ⫽4,329.63, p⬍.001, and the size

of the Kaiser–Meyer–Olkin measure of sampling adequacy (.89)

revealed that the set of IES items had adequate common variance

for factor analysis (Tabachnick & Fidell, 1996). To evaluate the

structure of the IES, I used a common factor analysis with prin-

cipal axis factoring and direct oblimin rotation. I chose this type of

rotation because I expected the factors to be correlated. The delta

weight was specified to be zero; this value permits a moderate

correlation between the factors. The number of factors was deter-

mined by factor eigenvalues above 1.0 and a noticeable change in

the slopes within the scree plot (Tabachnick & Fidell, 1996). The

rotated factor matrix was examined to pinpoint items that loaded

on these factors. Criteria for factor loadings included item values

greater than or equal to .40 on the primary factor and values less

than or equal to .30 on other factors. Although a common guideline

is to interpret loadings of .32 or higher (Tabachnick & Fidell,

1996), the minimum loading cutoff was set at .40 in order to

maximize confidence in the factors derived from the solution.

Six factors had eigenvalues greater than 1.0. Initial eigenvalues

and percentage of variance accounted for by each of these factors

were 7.14 and 27.45% for Factor 1, 3.64 and 14.00% for Factor 2,

1.83 and 7.04% for Factor 3, 1.40 and 4.98% for Factor 4, 1.27 and

4.54% for Factor 5, and 1.05 and 3.76% for Factor 6. Together,

they accounted for 61.77% of the variance. After inspecting the

scree plot, I observed a notable difference in the slopes of the first

three factors from those of subsequent factors. Therefore, the

factor solution of only these three factors was examined.

Three items that had either factor loadings less than .40 or

cross-loadings greater than or equal to .30 were eliminated. This

procedure resulted in 25 items, with Factor 1 containing 11 items,

Factor 2 containing 8 items and Factor 3 containing 6 items. Next,

these 25 items were factor analyzed with a principal-axis factor

analysis, three factors, and a direct oblimin rotation (

␦

⫽0). All

items loaded greater than .40 on their respective factor and less

than .30 on any other factor. This three-factor solution accounted

for 49.84% of the variance of the data. As expected, all items

loaded on the intuitive eating factor for which they were written.

Table 2 demonstrates these factor loadings.

The first factor (eigenvalue ⫽7.02) accounted for 28.08% of

the variance; its factor loadings ranged from .45 to .81. This

factor and subscale was labeled Unconditional Permission to

Eat. The second factor (eigenvalue ⫽3.61) accounted for

14.44% of the variance; its factor loadings ranged from .41 to

.79. This factor and subscale was labeled Eating for Physical

Table 1

Means, Standard Deviations, and Correlations Among the Measures of Study 1 (N ⫽391)

Measure MSD 123456789

1. IES: Total 3.36 0.56 — .84** .70** .62** ⫺.66** ⫺.53** ⫺.46** ⫺.52** ⫺.47**

2. IES: Unconditional Permission 3.41 0.82 .84** — .27** .29** ⫺.72** ⫺.44** ⫺.31** ⫺.41** ⫺.41**

3. IES: Eating for Physical Reasons 3.13 0.78 .72** .29** — .41** ⫺.23** ⫺.31** ⫺.41** ⫺.37** ⫺.29**

4. IES: Reliance on Hunger/Satiety Cues 3.59 0.53 .62** .32** .43** —

5. Eating Attitudes Test-26 2.41 0.68 ⫺.69** ⫺.76** ⫺.24** ⫺.27** —

6. EDI-2: Body Dissatisfaction 3.88 1.24 ⫺.56** ⫺.48** ⫺.33** ⫺.35** .54** —

7. EDI-2: Interoceptive Awareness 2.56 0.73 ⫺.49** ⫺.36** ⫺.39** ⫺.28** .53** .38** —

8. PSPS: Pressure for Thinness 2.30 0.90 ⫺.55** ⫺.46** ⫺.37** ⫺.28** .53** .59** .45** —

9. SATAQ: Internalization 3.39 0.95 ⫺.50** ⫺.44** ⫺.31** ⫺.23** .58** .57** .36** .47** —

Note. Values below the diagonal are for the 25-item Intuitive Eating Scale (IES) (M⫽3.36, SD ⫽0.56), the 11-item Unconditional Permission to Eat

subscale (M⫽3.41, SD ⫽0.82), and the 8-item Eating for Physical Reasons subscale (M⫽3.13, SD ⫽0.78); values above the diagonal are for the revised

21-item IES (

␣

⫽.88; M⫽3.31, SD ⫽0.58), the 9-item Unconditional Permission to Eat subscale (

␣

⫽.89; M⫽3.36, SD ⫽0.87), and the 6-item Eating

for Physical Reasons subscale (

␣

⫽.86; M⫽2.96, SD ⫽0.85). EDI-2 ⫽Eating Disorder Inventory—2; PSPS ⫽Perceived Sociocultural Pressures Scale;

SATAQ ⫽Sociocultural Attitudes Toward Appearance Questionnaire.

** p⬍.001.

229

INTUITIVE EATING SCALE

Rather Than Emotional Reasons. The third factor (eigenvalue ⫽

1.83) accounted for 7.32% of the variance; its factor loadings

ranged from .41 to .58. This factor and subscale was labeled

Reliance on Internal Hunger/Satiety Cues.

Internal consistency reliability evidence for the IES. The in-

ternal consistency reliabilities (alphas) for scores on the total IES

and subscales were .89 for the total 25-item IES, .89 for the

Unconditional Permission to Eat subscale, .86 for the Eating for

Physical Rather Than Emotional Reasons subscale, and .72 for the

Reliance on Internal Hunger/Satiety Cues subscale. For each sub-

scale, corrected item-total correlations were all above .30.

Validity evidence for the IES. Correlations between the IES

and the other study measures are presented in Table 1 below the

diagonal. Correlations around .10 were considered small or negli-

gible, correlations around .30 were considered moderate, and cor-

relations at or above .50 were considered large (Cohen, 1992;

Walsh & Betz, 2001). However, it should be noted that correlation

coefficients do not reflect the practical significance of a relation-

ship (Rosenthal, 1990); thus, it is important to consider this fact

when interpreting the strengths of these coefficients. First, the

hypothesis that the IES was related in a negative direction to eating

disorder symptomatology was explored. As expected, the total IES

and the Unconditional Permission to Eat subscale were strongly

related in a negative direction to eating disorder symptomatology;

the Eating for Physical Rather Than Emotional Reasons and the

Reliance on Internal Hunger/Satiety Cues subscales were slightly

to moderately related in a negative direction to eating disorder

symptomatology.

Second, the hypothesis that the IES would be negatively related

to body dissatisfaction, poor interoceptive awareness, pressure for

thinness, and internalization of the thin-ideal stereotype was in-

vestigated. As predicted, the total IES was strongly related in a

negative direction to these variables. For the IES subscales, Un-

conditional Permission to Eat was moderately to strongly related in

a negative direction to these variables, Eating for Physical Rather

Than Emotional Reasons was moderately related in a negative

direction to these variables, and Reliance on Internal Hunger/

Satiety Cues was slightly to moderately related in a negative

direction to these variables. Collectively, these findings provide

initial support for the IES’s construct validity.

Table 2

Item Factor Loadings and Corrected Item-Total Correlations for Each Intuitive Eating Scale

Factor Obtained From Analyzing the Data of Study 1 (N ⫽391)

Factor and item

Factor

Item-total: r123

Factor 1: Unconditional Permission to Eat (

␣

⫽.89)

1. I try to avoid certain foods... .70 ⫺.11 ⫺.05 .61

4. If I am craving a certain food... .45 ⫺.21 .07 .37

5. I follow eating rules... .65 ⫺.08 .06 .62

13. I try to ignore... .51 ⫺.06 .18 .53

14. I get mad at myself... .79 .13 ⫺.04 .77

19. I have forbidden foods... .65 ⫺.02 .05 .63

23. I feel guilty... .81 .15 ⫺.02 .79

25. I think of a certain food... .75 .01 ⫺.12 .67

26. I don’t trust myself... .55 .28 .02 .59

27. I don’t keep certain foods... .71 .28 ⫺.03 .72

28. I use caffeine... .45 .11 .09 .48

Factor 2: Eating for Physical Rather Than Emotional Reasons (

␣

⫽.86)

2. I stop eating... ⫺.13 .58 .09 .54

3. I find myself eating when I’m emotional... .17 .66 ⫺.03 .62

8. I find myself eating when I am bored... ⫺.11 .68 ⫺.03 .60

9. I eat everything on my plate... .05 .41 .14 .42

15. I find myself eating when I am lonely... .11 .67 .07 .68

21. I use food to help me soothe... .07 .79 ⫺.05 .72

22. I find myself eating when I am stressed... .13 .78 ⫺.02 .73

24. I find myself still eating... .05 .57 .09 .58

Factor 3: Reliance on Internal Hunger/Satiety Cues (

␣

⫽.72)

11. I can tell when I’m slightly full. .02 ⫺.07 .55 .37

12. I can tell when I’m slightly hungry. .01 ⫺.11 .57 .36

16. I trust my body to tell me when... .01 .22 .51 .57

17. I trust my body to tell me what... .01 .14 .41 .41

18. I trust my body to tell me how much... .09 .28 .49 .58

20. When I’m eating... .02 .04 .58 .43

230 TYLKA

Study 2: Confirmatory Factor Analysis and Additional

Construct Validity

It is recommended that a scale’s factor structure be analyzed

with confirmatory factor analysis in order to determine the overall

fit of the data to the scale model and whether items load on their

hypothesized latent factor(s) (Tabachnick & Fidell, 1996). Further-

more, when latent factors are expected to be related and connected

to a higher order factor, as in the IES’s structure, confirmatory

factor analysis allows for the estimation of the relationships be-

tween the latent factors and the determination of whether the latent

factors load on a higher order factor. Therefore, via a second-order

confirmatory factor analysis, it was hypothesized that the IES’s

items would load on their respective latent factors, that the latent

factors would be related, that the latent factors would load on a

higher order intuitive eating factor, and that the overall model

would provide an acceptable fit to the data.

Determining whether the IES would yield additional construct

validity evidence was also an aim of this study. It has been asserted

that intuitive eating is associated with psychological well-being

(Tribole & Resch, 1995). Specifically, dieting is associated with

(a) decreased self-esteem among women because it encourages

them to equate their self-worth with their body size (Fredrickson &

Roberts, 1997) and (b) feelings of deprivation and negative affect

that can decrease satisfaction with life and optimism (Polivy &

Herman, 1999). It has been proposed that because they are not

experiencing dieting-instigated negative affect, people with intui-

tive eating have higher self-esteem, satisfaction with life, opti-

mism, and use of adaptive coping strategies (Tribole & Resch,

1995). Given the above theory and research, it is hypothesized that

IES scores would be related in a positive direction to self-esteem,

optimism, proactive coping (i.e., efforts to develop resources that

lead to challenging goals and personal growth; Greenglass,

Schwarzer, & Taubert, 1999), and satisfaction with life. Last, IES

scores should not be related to response style. Therefore, it was

hypothesized that the IES would yield nonsignificant relations to

impression management, a biased form of responding that reflects

the tendency to give inflated self-descriptions to an audience.

Method

Participants and procedure. Participants read a description of the

study and enrolled via the psychology department Web site. The study was

described as an investigation of the relationships between eating habits,

personality characteristics, and response styles. After participants signed

the informed consent form and were guaranteed anonymity, they com-

pleted the measures, which were counterbalanced, in a classroom used as

a research laboratory. They received course credit for their involvement.

Responses from 7 women who did not answer 90% or more of any given

measure were not included in the data set. The final data set included

responses from 476 women from general psychology classes at a large

Midwestern university. Participants ranged in age from 17 to 50 (M⫽

19.70, SD ⫽4.50), and most (86.2%) identified as Caucasian American,

followed in frequency by Asian American (5.3%), African American

(3.9%), Latina (2.1%), and multiracial (2.4%). A large majority of the

participants were freshmen (72.6%); of the remaining participants, 14.6%

were sophomores, 4.6% were juniors, 7.7% were seniors, and 2 participants

(0.4%) did not indicate their college rank. Many women described them-

selves as upper middle class (47.9%) and middle class (41.6%), whereas

fewer women endorsed working-class (7.6%) and upper class (1.9%)

labels.

Measures. The 25-item IES, discussed in Study 1, was used in Study

2. Internal reliability (coefficient alpha) for the total IES was .87, internal

reliability for the Unconditional Permission to Eat subscale was .88,

internal reliability for the Eating for Physical Rather Than Emotional

Reasons subscale was .86, and internal reliability for the Reliance on

Internal Hunger/Satiety Cues subscale was .76.

In addition to the IES, the following scales were given to participants.

These scales frequently have been used to assess their respective constructs

(e.g., Lopez & Snyder, 2003; Paulhus, 1994).

Self-esteem was assessed by the Rosenberg Self-Esteem Scale (Rosen-

berg, 1965). It contains 10 items (e.g., “I feel that I have a number of good

qualities”) rated on a 4-point scale ranging from 1 (strongly disagree)to4

(strongly agree). Items are averaged, and higher scores reflect greater

self-esteem. Among college women, an alpha of .93 has been reported for

its scores (Tylka & Subich, 2004). For college students, the stability of its

scores over a 2-week period was .85, and it was related (r⫽.59) to another

measure of self-esteem (Robinson & Shaver, 1973). For the present study,

alpha was .90 for its scores.

The Life Orientation Test—Revised (Scheier, Carver, & Bridges, 1994)

contains six items (e.g., “In uncertain times, I usually expect the best”) that

assess generalized optimism and four filler items that are rated on a 4-point

scale ranging from 1 (strongly disagree)to4(strongly agree). In order to

obtain a total score, the six nonfiller items are averaged; higher scores

indicate a greater optimistic life orientation. Among a sample of college

students, the internal consistency reliability for its scores was .82, and it

was related to self-esteem (r⫽.54), self-mastery (r⫽.55), trait anxiety (r

⫽⫺.59), and neuroticism (r⫽⫺.50), supporting its construct validity

(Scheier et al., 1994). For the present study, alpha was .86 for its scores.

The Proactive Coping subscale from the Proactive Coping Inventory

(Greenglass et al., 1999) was used to assess women’s tendency to engage

in proactive coping. It contains 14 items (e.g., “I always try to find a way

to work around obstacles; nothing really stops me”) rated on a 5-point scale

ranging from 1 (strongly disagree)to5(strongly agree). Items are aver-

aged; higher scores reflect greater use of proactive coping. Among Cana-

dian college students, an alpha of .85 for its scores has been reported, and

evidence for its construct validity was garnered because it was related to

proactive attitudes (r⫽.70) and generalized self-efficacy (r⫽.70)

(Greenglass et al., 1999). For the present study, alpha was .89 for its scores.

The Satisfaction With Life Scale (Diener, Emmons, Larsen, & Griffin,

1985) was used to assess women’s global life satisfaction. Its five items

(e.g., “In most ways my life is close to ideal”) are rated along a 7-point

scale ranging from 1 (strongly disagree)to7(strongly agree). Items are

averaged to obtain a total score; higher scores indicate greater satisfaction

with life. Among college students, alpha was .87 for its scores, the stability

of its scores over a 2-month period was .82, and it was related to positive

affect (r⫽.50) and self-esteem (r⫽.54), supporting its construct validity

(Diener et al., 1985). For the present study, alpha was .91 for its scores.

The Impression Management subscale of the Balanced Inventory of

Desirable Responding Version 6 (Paulhus, 1994) assesses the tendency to

present inflated self-descriptions to others. Its 20 items (e.g., “I never

swear”) are rated on a 5-point scale ranging from 1 (not at all true of me)

to5(very true of me) and averaged. Higher scores indicate greater impres-

sion management. Among college students, alpha was .86 for its scores,

and the stability of its scores over a 5-week period was .77 (Paulhus, 1994).

For the present study, alpha was .77 for its scores.

Results and Discussion

The sample was large enough to perform confirmatory factor

analysis on the IES items (Bentler, 1990) in that 59 parameters

231

INTUITIVE EATING SCALE

were estimated. First, data were examined to ensure that the IES

items’ distributions were in accordance with the statistical assump-

tions of confirmatory factor analysis. No substantial violation was

indicated within the data; therefore, no items were transformed.

Mplus Version 2.12 (Muthe´n & Muthe´n, 2001) with maximum

likelihood estimation was used to perform the confirmatory factor

analysis. The 25 IES items served as indicators of their respective

first-order latent factor (i.e., Unconditional Permission to Eat,

Eating for Physical Rather Than Emotional Reasons, Reliance on

Internal Hunger/Satiety Cues). Relations between the three hy-

pothesized latent factors were estimated, and a second-order (i.e.,

Intuitive Eating) latent factor was estimated from the first-order

factors. The adequacy of fit was determined by the four indices

calculated by Mplus and recommended by Hu and Bentler (1999):

the comparative fit index (CFI), the Tucker–Lewis Index (TLI)—

also known as the nonnormed fit index—the standardized root-

mean-square residual (SRMR), and the root-mean-square error of

approximation (RMSEA). The fit statistics ranged from poor (i.e.,

CFI ⫽.83, TLI ⫽.81) to fair (RMSEA ⫽.09, SRMR ⫽.09) as

determined by criteria for model fit adequacy (Browne & Cudeck,

1993; Hu & Bentler, 1999). Contrary to hypotheses, this model did

not provide a good overall fit to the data.

Consequently, factor loadings for this model were evaluated to

determine whether certain items did not load strongly on their

hypothesized latent factor and whether the deletion of such items

would enhance the fit of the model to the data. Other researchers

(e.g., Phillips, Szymanski, Ozegovic, & Briggs-Phillips, 2004)

have discarded items that load below .45 on their hypothesized

factor; thus, I decided to delete four items (Items 9, 13, 24, and 28)

that loaded at .37, .44, .44, and .42, respectively, on their hypoth-

esized latent factor. A second confirmatory factor analysis then

was conducted with the remaining 21 items as indicators of their

respective first-order latent factor. Similar to the first model,

relations between the first-order factors were estimated, and a

second-order intuitive eating latent factor was estimated from the

first-order factors. This revised model provided an adequate fit to

the data (Browne & Cudeck, 1993; Hu & Bentler, 1999) because

all fit statistics were acceptable (i.e., CFI ⫽.91, TLI ⫽.90,

RMSEA ⫽.08, SRMR ⫽.07). Figure 1 displays the item and

factor loadings from this analysis. All items loaded significantly on

their respective first-order latent factors, and each first-order factor

loaded significantly on the second-order intuitive eating factor;

thus, the hypothesized latent factors are internally consistent and

exist empirically. The first-order factors were moderately related

to one another. Consequently, the remaining analyses were con-

ducted by using this 21-item version of the IES, which is presented

in Appendix A.

Validity evidence for the IES. First, the hypothesis that the IES

would be related in a positive direction to self-esteem, optimism,

proactive coping, and satisfaction with life was explored. As

predicted, IES total scores were moderately to strongly related to

self-esteem and satisfaction with life and moderately related to

optimism and proactive coping. The Unconditional Permission to

Eat subscale was moderately related to self-esteem and satisfaction

with life; however, it was only negligibly related to optimism and

unrelated to proactive coping. The Eating for Physical Rather Than

Emotional Reasons subscale was moderately to strongly related to

self-esteem and moderately related to optimism, proactive coping,

and satisfaction with life. Last, the Reliance on Internal Hunger/

Satiety Cues was moderately to strongly related to satisfaction

with life and was moderately related to self-esteem, optimism, and

proactive coping. As demonstrated in Table 3, all relationships

were in a positive direction, indicating that higher IES scores are

associated with higher levels of psychological health. Collectively,

these findings lend additional support for the IES’s construct

validity.

Last, the hypothesis that intuitive eating would be either unre-

lated or negligibly related to impression management was ex-

plored. As expected, this hypothesis was supported because im-

pression management was not related to the total IES, the

Unconditional Permission to Eat subscale, or the Eating for Phys-

ical Rather Than Emotional Reasons subscale. It was only negli-

gibly related to the Reliance on Internal Hunger/Satiety Cues

subscale. Table 3 presents these relationships.

Internal consistency reliability evidence for the revised IES

scores. Alphas were recalculated for the total IES and the first

two subscales, which were affected by the deletion of the four IES

items. Alphas were .85 for the total IES scores, .87 for the

Unconditional Permission to Eat subscale scores, and .85 for the

Eating for Physical Rather Than Emotional Reasons subscale

scores.

Reanalyzing the data of Study 1 by using the revised 21-item

IES. For Study 1, the IES’s means, standard deviations, alpha

levels, and intercorrelations with the other Study 1 measures were

recalculated by using the 21-item version. This information is

presented in Table 1. As indicated, similar values were obtained

for the 21-item version.

Study 3: The Relation of the IES to Body Mass:

Additional Construct Validity

Because intuitive eaters allow their internal hunger cues to guide

their food intake and follow their internal satiety cues to determine

when to stop eating, they are more likely to weigh an amount that

is ideal for their body type (Polivy & Herman, 1992; Tribole &

Resch, 1995). Conversely, approximately 95% of those who diet

and lose weight will regain the weight within a few years, and

many of these individuals will gain more weight than they origi-

nally lost (Heatherton, Mahamedi, Striepe, Field, & Keel, 1997).

Dieting also has been associated with increased food preoccupa-

tion, binge eating, and eating in the absence of hunger (Birch et al.,

2003; Keys et al., 1950; Polivy & Herman, 1999). Indeed, indi-

viduals who do not diet were found to be less likely to eat in

response to emotional fluctuations and situational factors (e.g.,

visual and olfactory food cues) than those who diet (Carper et al.,

2000; Fedoroff et al., 1997; Kahan et al., 2003). Overall, these

findings indicate that individuals who eat intuitively are less likely

to engage in behaviors that may lead to weight gain (e.g., eating in

the absence of hunger, eating in response to emotional fluctuations

and situational factors, binge eating) than people who diet. Even

though some average and below average weight women may

practice restrained eating without engaging in these behaviors,

women who eat intuitively should have, on average, lower body

mass than those who restrain their eating, because dieting is an

ineffective means of weight loss for most individuals (Heatherton

232 TYLKA

et al., 1997). Therefore, it was hypothesized that IES scores would

be related in a negative direction to women’s body mass; body

mass index (BMI) was examined in lieu of weight because it

controls for height.

Method

Participants and procedure. Participants were enrolled in introductory

psychology courses and learned about the study through the psychology

department’s organized research program. They were informed that the

study was an investigation of their eating habits. Participants were in-

formed that their responses would remain anonymous. After providing

their consent, participants completed the survey in a classroom used as a

research lab. They received credit that was applied toward their class grade.

Responses from 3 women who did not answer at least 90% of the IES or

did not report their weight or height were not included in the data set.

Responses from 199 women (Mage ⫽18.92 years, SD ⫽3.25, range ⫽

17–55) from a large Midwestern university were entered into the data set.

Women identified as Caucasian American (75.4%), African American

(13.1%), Asian American (4.0%), Latina (2.0%), international (3.5%),

multiracial (1.5%), and Native American (0.5%). Most participants were

freshmen (77.4%), whereas 16.6% were sophomores, 3.5% were juniors,

1.0% were seniors, and 1.5% did not report their college rank. Most women

described themselves as middle class (43.7%) and upper middle class

(33.2%); fewer women described themselves as working class (8.0%) and

upper class (4.0%).

Measures. The 21-item IES was used. Participants also were asked to

report their current weight and height; this information was garnered to

estimate participants’ BMI. Because college women’s actual BMI has been

found to be highly correlated with the BMI calculated from their self-

ratings of weight and height (r⫽.98, p⬍.001; Tylka & Subich, 1999),

I decided to assess only self-ratings of these variables. The average BMI

for the current sample was 23.50 (SD ⫽3.90; range ⫽17.47–34.94); this

average score is within the normal range (i.e., 22–25) recommended for

women (Kuczmarski & Flegal, 2000). Average IES scores were 3.09

(SD ⫽0.57) for the total IES, 3.08 (SD ⫽0.88) for the Unconditional

Figure 1. Factor loadings of the Intuitive Eating Scale (IES) garnered via second-order confirmatory factor

analysis of the revised 21-item IES data of Study 2 (N⫽476). *p⬍.01.

233

INTUITIVE EATING SCALE

Permission to Eat subscale, 2.65 (SD ⫽0.88) for the Eating for Physical

Rather Than Emotional Reasons subscale, and 3.56 (SD ⫽0.63) for the

Reliance on Internal Hunger/Satiety Cues subscale. Alphas were .85 for the

total IES scores, .87 for the Unconditional Permission to Eat subscale

scores, .85 for the Eating for Physical Rather Than Emotional Reasons

subscale scores, and .78 for the Reliance on Internal Hunger/Satiety Cues

subscale scores.

Results and Discussion

As hypothesized, IES scores were negatively related to BMI.

The relationships between BMI and IES scores were –.28 ( p⬍

.001) for the total IES, –.21 for the Unconditional Permission to

Eat subscale ( p⬍.01), –.17 for the Eating for Physical Rather

Than Emotional Reasons subscale ( p⬍.05), and –.20 for the

Reliance on Internal Hunger/Satiety Cues subscale ( p⬍.01).

These relationships are slight to moderate in size and are consistent

with theory and research (e.g., Kahan et al., 2003; Tribole &

Resch, 1995) suggesting that listening to body signals in deter-

mining what, when, and how much to eat is associated with lower

body mass. These findings support the construct validity of the IES

because they are consistent with the tenets of intuitive eating.

Study 4: Test–Retest Reliability Estimates

When investigating the psychometric properties of a scale that

claims to measure a stable construct, it is important to examine

whether the scale yields consistent scores over a given time period

(Walsh & Betz, 2001). Because intuitive eating has been proposed

to be fairly consistent over time (Tribole & Resch, 1995), it was

hypothesized that the total IES and subscale scores would be stable

over a 3-week period, which would support the test–retest reliabil-

ity of its scores.

Method

Women at the regional campus were recruited via verbal announcements

of the experiment given in their psychology classes, and women at the main

campus learned of the study through the psychology department’s orga-

nized research program. They were informed that the study was an inves-

tigation of their eating habits. For each administration, participants were

instructed to write a code consisting of the first two letters of their mother’s

maiden name and the last two digits of their house or apartment number on

their questionnaire. This code permitted the matching of their initial and

follow-up responses. After participants were assured of the anonymity of

their responses and provided their consent, they completed the IES in a

classroom used as a research lab. They completed the IES for the second

time exactly 3 weeks later and received extra credit that was applied toward

their grade.

Responses from 17 women who did not complete at least 90% of the IES

during the first and second administration (16 of these women did not

complete the IES at the second administration) were not entered into the

data set. The final data set included responses from 194 women (mean

age ⫽22.07 years, SD ⫽7.38, range ⫽17–55) enrolled in general and

upper level psychology classes at a regional campus (n⫽67) and the main

campus (n⫽127) of a large Midwestern university. Women identified as

Caucasian American (94.3%), multiracial (2.6%), African American

(2.1%), Latina (0.5%), and Native American (0.5%). Most participants

were freshmen (54.1%), whereas 15.5% were sophomores, 9.8% were

juniors, 19.1% were seniors, and 3 participants (1.5%) did not report their

college rank. Many women described themselves as middle class (60.8%);

fewer women endorsed upper middle class (21.1%), working-class

(17.5%), and upper class (0.5%) labels.

For the first and second administration, respectively, mean scores and

alphas were 3.41 (SD ⫽.53;

␣

⫽.86) and 3.43 (SD ⫽.54;

␣

⫽.89) for

the total IES, 3.50 (SD ⫽.82;

␣

⫽.89) and 3.49 (SD ⫽.80;

␣

⫽.91) for

the Unconditional Permission to Eat subscale, 3.03 (SD ⫽.85;

␣

⫽.87)

and 3.01 (SD ⫽.84;

␣

⫽.89) for the Eating for Physical Rather Than

Emotional Reasons subscale, and 3.67 (SD ⫽.53;

␣

⫽.72) and 3.74 (SD ⫽

.51;

␣

⫽.78) for the Reliance on Internal Hunger/Satiety Cues subscale.

Mean IES total and subscale scores did not differ between the regional and

main campus settings, nor did they differ between general psychology and

upper level psychology students (all ps⬎.05).

Results and Discussion

As hypothesized, the stability of the IES over a 3-week period

was supported. The relationship between the first and second

administration was .90 for the total IES, .88 for the Unconditional

Permission to Eat subscale, .88 for the Eating for Physical Rather

Than Emotional Reasons subscale, and .74 for the Reliance on

Internal Hunger/Satiety Cues subscale (all ps⬍.001).

Table 3

Means, Standard Deviations, and Correlations Among the Measures of Study 2 (N ⫽476)

Measures MSD 123456789

1. IES 3.24 0.54 —

2. IES: Unconditional Permission 3.19 0.83 .81** —

3. IES: Eating for Physical Reasons 2.91 0.82 .66** .18** —

4. IES: Reliance on Hunger/Satiety Cues 2.72 0.41 .60** .24** .33** —

5. Rosenberg Self-Esteem Scale 3.21 0.50 .44** .28** .36** .35** —

6. LOT–R (Optimism) 2.90 0.50 .29** .14* .24** .31** .73** —

7. PCI: Proactive Coping 3.77 0.52 .29** .10 .27** .34** .63** .67** —

8. Satisfaction With Life Scale 5.08 1.14 .41** .26** .26** .38** .68** .61** .54** —

9. BIDR-6: Impression Management 2.82 0.51 .12 .07 .06 .16** .21** .15** .18** .10 —

Note. The IES is based on the 21-item scale, Unconditional Permission to Eat is based on the 9-item subscale, and Eating for Physical Reasons is based

on the 6-item subscale. IES ⫽Intuitive Eating Scale; LOT–R ⫽Life Orientation Test—Revised; PCI ⫽Proactive Coping Inventory; BIDR-6 ⫽Balanced

Inventory of Desirable Responding— 6.

*p⬍.01. ** p⬍.001.

234 TYLKA

General Discussion

Intuitive eating is characterized by eating according to internal

hunger and satiety signals; it is not akin to binge eating, where

individuals allow themselves to eat but cannot control their eating.

In four studies, the development of a measure of intuitive eating

(the IES) was discussed, its factor structure was investigated, and

reliability and validity evidence for its scores was garnered among

college women. An exploration of the IES’s factor structure in

Study 1 revealed that its 25 items formed three conceptually

meaningful factors, which accounted for approximately 50% of its

variance. However, findings from a second-order confirmatory

factor analysis conducted in Study 2 suggested that 4 of these

original IES items did not load strongly on their respective latent

factors and were subsequently deleted from the scale. The resultant

21-item scale provided a good fit to the data and consequently was

used to obtain the findings presented in this discussion. The first

factor–subscale, Unconditional Permission to Eat, consisted of 9

items that assessed the willingness to eat when physiologically

hungry and what food is desired at the moment. The second

factor–subscale, Eating for Physical Rather Than Emotional Rea-

sons, consisted of 6 items that reflected the tendency to eat to

satisfy an internal hunger drive rather than to cope with emotional

fluctuations and/or distress. The third factor–subscale, Reliance on

Internal Hunger/Satiety Cues, contained 6 items that reflected the

degree of awareness of internal hunger and satiety signals and the

ability of these signals to guide eating behavior.

The second-order confirmatory factor analysis further revealed

that the hypothesized latent factors (formed from the subscale

items) were moderately related to one another and loaded on a

higher order latent factor. These findings provide empirical sup-

port for the assertion that IES factors are theoretically distinct

albeit related components of the broader intuitive eating construct.

Moreover, the total IES and its subscale scores were found to be

internally consistent in all studies and stable over a 3-week period

in Study 4.

This study also obtained construct validity evidence for the total

IES and its subscales. Claims made by scholars (e.g., Costanzo et

al., 2001; Mills et al., 2002; Polivy & Herman, 1999; Tribole &

Resch, 1995) that intuitive eating should be negatively related to

eating disorder symptomatology, body dissatisfaction, poor intero-

ceptive awareness, pressure for thinness, and internalization of the

thin-ideal stereotype were supported in the present study because

Study 1 demonstrated that the total IES and the IES subscales were

negatively related to these variables. Consistent with theory and

research on intuitive eating behaviors (e.g., Costanzo et al., 2001;

Polivy & Herman, 1999; Tribole & Resch, 1995), Study 2 revealed

that IES total and subscale scores were positively related to mea-

sures of self-esteem, optimism, proactive coping, and satisfaction

with life (with the exception of the nonsignificant relation between

Unconditional Permission to Eat and proactive coping), further

supporting the IES’s connection to psychological well-being.

Study 3 found that IES total and subscale scores were negatively

related to BMI, which upholds the connection documented be-

tween dieting and elevated weight (Birch et al., 2003; Carper et al.,

2000; Heatherton et al., 1997). In addition, the IES total and

subscale scores were either not related or negligibly related to an

impression management response style, which supports the dis-

criminant validity of the IES.

The findings that women with higher levels of intuitive eating

are more satisfied with their bodies and perceive less pressure to be

thin may be explained, at least in part, by the negative association

between IES scores and body mass (i.e., they more closely resem-

ble the societal thin-ideal body type). Yet, these women were more

likely to reject the societal stereotype that thinness is their ideal

body type, indicating that they are less likely to base their self-

worth on being thin.

Furthermore, women scoring higher on the Eating for Physical

Rather Than Emotional Reasons subscale may not use food to cope

with their emotional distress because they experience less of this

distress, whereas women who score lower on this subscale may

experience more emotional distress and eat as a coping, albeit

maladaptive, strategy. However, women scoring higher on this

subscale also scored higher on proactive coping, indicating that

they reported using more constructive strategies to deal with their

emotional distress.

Collectively, these findings support the IES and the adaptive

properties of intuitive eating. Given that the IES is relatively brief

and easy to administer and score, it would be useful for both

researchers and clinicians who work with women in a variety of

venues. The total IES and subscale mean values, standard devia-

tions, and skewness and kurtosis values were calculated for the

total sample of women (N⫽1,260) and for the various ethnic

groups, with the data collapsed across all four studies. These

values are presented in Appendix B.

Limitations and Future Research

Most of the participants in the present study were young-adult,

Caucasian, middle- to upper middle-class first-year psychology

students. It is important to determine whether the IES yields

reliable and valid scores with other samples of women (e.g.,

community women or women of color). Also, future research

endeavors could be aimed at demonstrating whether its psycho-

metric properties are upheld with men. Scores on the IES may be

similar for women and men because both women and men are

socialized to adopt an external orientation to food intake: whereas

women are socialized to lose weight, men are socialized to gain

muscle mass (Vartanian, Giant, & Passino, 2001). Men may in-

ternalize the cultural pressure to restrict their eating to high-protein

foods and eat these foods when they are not hungry in order to gain

weight, which will negatively impact their IES scores.

Another limitation is that many of the first and second subscale

items were worded in the direction of maladaptive eating (e.g., “I

find myself eating when I am stressed out, even when I’m not

physically hungry”) and then reverse scored. Reverse scoring these

negative items may not yield the most valid possible measure of

the converse construct. Amending many of these items in the

direction of adaptive eating (e.g., “I am able to cope with my stress

without turning to food for comfort”) and determining the psycho-

metric properties of this revised scale is recommended.

The present study’s exclusive use of self-report methodology

also is somewhat limiting because it relies on participants’

accurate reporting of their current level of functioning. In

235

INTUITIVE EATING SCALE

particular, participants in Study 3 could have either distorted

their weight and height or not known and estimated this infor-

mation. Furthermore, women’s perceptions of their eating hab-

its may or may not be an accurate portrayal of reality. Thus,

another avenue for research could be to determine whether

self-reported eating practices on the IES are strongly correlated

with actual eating behaviors.

Research could focus on determining the best way to help

individuals who have developed an external orientation to food

intake (e.g., those who chronically diet) work toward eating ac-

cording to their internal hunger and satiety signals. These individ-

uals’ internal signals have been weakened as a result of being

ignored, numbed, and replaced with external cues as to when,

what, and how much to eat; consequently, these individuals may

be unable to tell when they are hungry or sated (Birch & Fisher,

2000; Garner, 1991). Because hunger and satiety signals are often

ambiguous, these individuals may interpret their emotional distress

as cues that they are physiologically hungry (Costanzo et al.,

2001).

The IES was related to eating disorder symptomatology, yet the

strength of this relation indicated that there was not a complete

overlap among these constructs (i.e., they shared 43.6% of the

variance). This finding provides preliminary support for the

uniqueness of intuitive eating, suggesting that it is negatively

related to, but more than the mere absence of, eating disorder

symptomatology. Nevertheless, it remains to be determined

whether intuitive eating predicts unique variance in psychological

health variables (e.g., self-esteem, subjective well-being) above

and beyond the variance accounted for by eating disorder

symptomatology.

Future studies might also examine whether adaptive environ-

mental influences (e.g., acceptance of body size from significant

others or caregivers’ use of noncoercive feeding strategies) predict

positive body image, which then predicts intuitive eating. This

model could provide insight as to how some women continue to

follow their internal physiological signals within a social milieu

that frequently pressures women to become thinner.

If future research confirms the validity and reliability of the IES

in samples of clients who present with concerns about food or

eating disorder symptoms, the IES may prove to be a valuable

clinical tool. For instance, the IES could be used to assess whether

clients increase adaptive eating behaviors and attitudes as a result

of treatment. In fact, a comprehensive treatment approach for

disordered eating should result in an increase in adaptive charac-

teristics as well as a reduction of maladaptive symptoms. Further-

more, given its attention to adaptive eating, the IES may better

predict relapse prevention than current eating disorder instruments,

such as the EAT-26 or Eating Disorder Inventory–3.

References

Bentler, P. M. (1990). Comparative fit indexes in structural models. Psy-

chological Bulletin, 107, 238 –246.

Birch, L. L., & Fisher, J. O. (2000). Mothers’ child-feeding practices

influence daughters’ eating and weight. American Journal of Clinical

Nutrition, 71, 1054 –1061.

Birch, L. L., Fisher, J. O., & Davison, K. K. (2003). Learning to overeat:

Maternal use of restrictive feeding practices promotes girls’ eating in the

absence of hunger. American Journal of Clinical Nutrition, 78, 215–220.

Birch, L. L., Johnson, S. L., Andresen, G., Petersen, J. C., & Schulte, M. C.

(1991). The variability of young children’s energy intake. New England

Journal of Medicine, 324, 232–235.

Brookings, J. B., & Wilson, J. F. (1994). Personality and family-

environment predictors of self-reported eating attitudes and behaviors.

Journal of Personality Assessment, 63, 313–326.

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model

fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation

models (pp. 136 –162). Newbury Park, CA: Sage.

Carper, J. L., Fisher, J. O., & Birch, L. L. (2000). Young girls’ emerging

dietary restraint and disinhibition are related to parental control in child

feeding. Appetite, 35, 121–129.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.

Costanzo, P. R., Reichmann, S. K., Friedman, K. E., & Musante, G. J.

(2001). The mediating effect of eating self-efficacy on the relationship

between emotional arousal and overeating in the treatment-seeking

obese. Eating Behaviors, 2, 363–368.

Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The

Satisfaction With Life Scale. Journal of Personality Assessment, 49,

71–75.

Faith, M. S., Scanlon, K. S., Birch, L. L., Francis, L. A., & Sherry, B.

(2004). Parent– child feeding strategies and their relationships to child

eating and weight status. Obesity Research, 12, 1711–1722.

Fedoroff, I. C., Polivy, J., & Herman, C. P. (1997). The effect of pre-

exposure to food cues on the eating behavior of restrained and unre-

strained eaters. Appetite, 28, 33– 47.

Fredrickson, B. L., & Roberts, T. A. (1997). Objectification theory: To-

ward understanding women’s lived experiences and mental health risks.

Psychology of Women Quarterly, 21, 173–206.

Garner, D. M. (1991). Eating Disorder Inventory–2 manual. Odessa, FL:

Psychological Assessment Resources.

Garner, D. M. (2004). Eating Disorder Inventory–3 manual. Odessa, FL:

Psychological Assessment Resources.

Garner, D. M., Olmsted, M. P., Bohr, Y., & Garfinkel, P. E. (1982). The

Eating Attitudes Test: Psychometric features and clinical correlates.

Psychological Medicine, 12, 871– 878.

Greenglass, E. R., Schwarzer, R., & Taubert, S. (1999). The Proactive

Coping Inventory (PCI): A multidimensional research instrument. Re-

trieved June 10, 2005, from http://www.psych.yorku.ca/greenglass

Heatherton, T. F., Mahamedi, F., Striepe, M., Field, A. E., & Keel, P.

(1997). A 10-year longitudinal study of body weight, dieting, and eating

disorder symptoms. Journal of Abnormal Psychology, 106, 117–125.

Heinberg, L. J., Thompson, J. K., & Stormer, S. (1995). Development and

validation of the Sociocultural Attitudes Towards Appearance Question-

naire. International Journal of Eating Disorders, 17, 81– 89.

Herman, C. P., & Polivy, J. (1983). A boundary model for the regulation

of eating. In A. J. Stunkard & E. Stellar (Eds.), Eating and its disorders

(pp. 141–156). New York: Raven.

Herman, C. P., & Polivy, J. (1988). Studies of eating in normal dieters. In

B. T. Walsh (Ed.), Eating behavior in eating disorders (pp. 95–112).

Washington, DC: American Psychiatric Association.

Herman, C. P., Polivy, J., Lank, C. N., & Heatherton, T. F. (1987).

Anxiety, hunger, and eating behavior. Journal of Abnormal Psychology,

96, 264 –269.

Hu, L., & Bentler, P. (1999). Cutoff criteria for fit indices in covariance

structure analysis: Conventional criteria versus new alternatives. Struc-

tural Equation Modeling, 6, 1–55.

Kahan, D., Polivy, J., & Herman, C. P. (2003). Conformity and dietary

disinhibition: A test of the ego-strength model of self-regulation. Inter-

national Journal of Eating Disorders, 33, 165–171.

Keys, A., Brozek, J., Henschel, A., Mickelsen, O., & Taylor, H. L. (1950).

236 TYLKA

The biology of human starvation (Vols. 1–2). Minneapolis, MN: Uni-

versity of Minnesota Press.

Kuczmarski, R. J., & Flegal, K. M. (2000). Criteria for definition of

overweight in transition: Background and recommendations for the

United States. American Journal of Clinical Nutrition, 72, 1075–1081.

Lopez, S. J., & Snyder, C. R. (Eds.). (2003). Positive psychological

assessment: A handbook of models and measures. Washington, DC:

American Psychological Association.

Mazzeo, S. E. (1999). Modification of an existing measure of body image

preoccupation and its relationship to disordered eating in female college

students. Journal of Counseling Psychology, 46, 42–50.

Mills, J. S., Polivy, J., Herman, C. P., & Tiggemann, M. (2002). Effects of

exposure to thin media images: Evidence of self-enhancement among

restrained eaters. Personality and Social Psychology Bulletin, 28, 1687–

1699.

Mintz, L. B., & Betz, N. E. (1988). Prevalence and correlates of eating

disordered behaviors among undergraduate women. Journal of Coun-

seling Psychology, 44, 463– 471.

Mintz, L. B., O’Halloran, M. S., Mulholland, A. M., & Schneider, P. A.

(1997). Questionnaire for eating disorder diagnoses: Reliability and

validity of operationalizing DSM–IV criteria into a self-report format.

Journal of Counseling Psychology, 44, 63–79.

Muthe´n, L. K., & Muthe´n, B. O. (2001). Mplus user’s guide (2nd ed.). Los

Angeles: Author.

Ogden, J. (2003). The psychology of eating: From healthy to disordered

behavior. Malden, MA: Blackwell.

Paulhus, D. L. (1994). Balanced Inventory of Desirable Responding ref-

erence manual for BIDR Version 6. Vancouver, British Columbia,

Canada: University of British Columbia.

Phillips, J. C., Szymanski, D. M., Ozegovic, J. J., & Briggs-Phillips, M.

(2004). Preliminary examination and measurement of the internship

research training environment. Journal of Counseling Psychology, 51,

240 –248.

Polivy, J., & Herman, C. P. (1987). The diagnosis and treatment of normal

eating. Journal of Consulting and Clinical Psychology, 55, 635– 644.

Polivy, J., & Herman, C. P. (1992). Undieting: A program to help people

stop dieting. International Journal of Eating Disorders, 11, 261–268.

Polivy, J., & Herman, C. P. (1999). Distress and eating: Why do dieters

overeat? International Journal of Eating Disorders, 26, 153–164.

Robinson, J. P., & Shaver, P. R. (1973). Measures of social psychological

attitudes (2nd ed.). Ann Arbor, MI: Institute for Social Research.

Rogers, C. R. (1964). Toward a modern approach to values: The valuing

process in the mature person. Journal of Abnormal and Social Psychol-

ogy, 68, 160 –167.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,

NJ: Princeton University Press.

Rosenthal, R. (1990). How are we doing in soft psychology? American

Psychologist, 45, 775–776.

Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing

optimism from neuroticism (and trait anxiety, self-mastery, and self-

esteem): A reevaluation of the Life Orientation Test. Journal of Person-

ality and Social Psychology, 67, 1063–1078.

Seligman, M. E., & Csikszentmihalyi, M. (2000). Positive psychology: An

introduction. American Psychologist, 55, 5–14.

Stice, E., Ziemba, C., Margolis, J., & Flick, P. (1996). The dual pathway

model differentiates bulimics, subclinical bulimics, and controls: Testing

the continuity hypothesis. Behavior Therapy, 27, 531–549.

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd

ed.). New York: HarperCollins.

Tribole, E., & Resch, E. (1995). Intuitive eating: A recovery book for the

chronic dieter. New York: St. Martin’s Press.

Tylka, T. L., & Subich, L. M. (1999). Exploring the construct validity of

the eating disorder continuum. Journal of Counseling Psychology, 46,

268 –276.

Tylka, T. L., & Subich, L. M. (2004). Examining a multidimensional

model of eating disorder symptomatology among college women. Jour-

nal of Counseling Psychology, 51, 314 –328.

Vartanian, L. R., Giant, C. L., & Passino, R. M. (2001). “Ally McBeal vs.

Arnold Schwarzenegger”: Comparing mass media, interpersonal feed-

back and gender as predictors of satisfaction with body thinness and

muscularity. Social Behavior and Personality, 29, 711–723.

Walsh, W. B., & Betz, N. E. (2001). Tests and assessment (4th ed.). Upper

Saddle River, NJ: Prentice Hall.

Wear, R. W., & Pratz, O. (1987). Test–retest reliability for the Eating

Disorder Inventory. International Journal of Eating Disorders, 6, 767–

769.

Woody, E., Costanzo, P., Leifer, H., & Conger, J. (1981). The effects of

taste and caloric perceptions on the eating behavior of restrained and

unrestrained subjects. Cognitive Therapy and Research, 5, 381–390.

(Appendixes follow)

237

INTUITIVE EATING SCALE

Appendix A

Intuitive Eating Scale (21 Items)

Directions for participants: For each item, please circle the answer that best characterizes your attitudes or behaviors.

1. I try to avoid certain foods high in fat, carbohydrates, or calories.

12345