The Spanish Journal of Psychology Copyright 2012 by The Spanish Journal of Psychology

2012, Vol. 15, No. 2, 801-816 ISSN 1138-7416

http://dx.doi.org/10.5209/rev_SJOP.2012.v15.n2.38892

Development and Validation of the Theory of Planned

Behavior Questionnaire in Physical Activity

Sonia Tirado González, Mª Carmen Neipp López, Yolanda Quiles Marcos,

and Jesús Rodríguez-Marín
Universidad Miguel Hernández (Spain)

The Theory of Planned Behavior (TPB) is one of the main theoretical models in the study of
the different variables, which influence in the practise of physical activity. The aim in this study
was to develop a questionnaire based on TPB in physical activity context providing evidence
for the validity of the obtained measures. The instrumental project included three independent
studies. The first study entailed the construction and qualitative assessment of the items. In the
second study, the analysis of factorial structure was performed by means of exploratory measures,
and it showed that the reliability of measures was adequate. The third study provided evidence
on the dimensionality of the scale. The confirmatory factorial analysis guaranteed the stability
of factorial structure proposed by the TPB and provided evidence for the internal validity of
the inventory. Moreover, this study provided evidence of its external validity.
Keywords: theory of planned behavior, physical exercise, instrumental study, validation.

La Teoría de la Conducta Planeada (TCP) es uno de los modelos teóricos más empleados para
conocer las diferentes variables que influyen en la realización de ejercicio físico. El objetivo en
este trabajo fue elaborar un cuestionario basado en la TCP en el contexto del ejercicio físico
y aportar evidencias sobre la validez de sus mediciones. El trabajo instrumental incluyó tres
estudios independientes. El primer estudio consistió en la elaboración y evaluación cualitativa
de los ítems. En el segundo estudio se analizó la estructura factorial mediante procedimientos
exploratorios y se mostró que la fiabilidad de las mediciones era adecuada. El tercer estudio
aportó evidencias sobre la dimensionalidad del inventario. El análisis factorial confirmatorio avaló
la estabilidad de la estructura factorial propuesta por la TCP, así como aportó evidencias de
validez interna del instrumento. También se aportan evidencias de su validez externa.
Palabras clave: teoría conducta planeada, ejercicio físico, estudio instrumental, validación.

Correspondence concerning this article should be addressed to Mª Carmen Neipp López. Departamento de Psicología de la Salud.
Universidad Miguel Hernández. Avda. de la Universidad s/n, Edificio Altamira. 03202 Elche – Alicante (Spain). E-mail: neipp@umh.es

801
802 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN
According to the latest information from the World
Health Organization (WHO), the majority of the adult
population is either sedentary or not sufficiently active
(WHO, 2002; 2004). In the case of the European Union,
two-thirds of the adult population do not receive the
recommended levels of physical activity. Spain would be
between the European countries whose populations exercise
less, where only 25% of adults can be considered sufficiently
active (Sjöström, Oja, Hagströmer, Smith, & Bauman, 2006;
WHO, 2006). Moreover, many people who begin exercise
programmes abandon them after short periods of time and
their inactivity returns within the first six months (Marcus
et al., 2000; Sniehotta, Scholz, & Schwarzer, 2005). It has
been sufficiently demonstrated that regular physical exercise
plays an important role in combating stress and as a factor
in improving a person’s quality of life. There is extensive
literature indicating that regular exercise is an effective
preventive strategy against non-communicable diseases like
obesity, type 2 diabetes, breast cancer, cardiovascular
diseases, hypertension, colon cancer and osteoporosis, among
others (Blair & Brodney, 1999; Taras, 2005). These non-
communicable diseases are those responsible for the largest
percentage of morbidity and mortality in developed countries
like Spain, and represent 60% of the world’s total mortality
and 47% of its morbidity (WHO, 2004). Nevertheless,
exercising regularly does not only have beneficial effects
upon physical health; vast research endorses the benefits of
physical exercise in different areas of psychological well-
being: it improves subjective health, moods and emotion
(Biddle, Fox, & Boutcher, 2000; Netz, Wu, Becker, &
Tenebaum, 2005), reduces clinical depression (Lawlor &
Hopker, 2001), diminishes levels of anxiety and stress
(Jiménez, Martínez, Miró, & Sánchez, 2008) and increases
self-esteem (McAuley, Mihalko, & Bane, 1997).
The high prevalence of physical inactivity shows the
need to design interventions that increase the practice of
physical exercise. In order to develop these interventions,
it is necessary to understand the processes that lead the
subject to initiate and maintain regular physical-sporting
Figure 1. Theory of Planned Behavior.
activity, including the motivations and different factors that
may influence their fulfilment, with the objective of
developing more appropriate intervention strategies. One
of the most-employed theoretical models to learn about the
factors that maintain or hinder the behavioral change
necessary to improve a person’s quality of life is the Theory
of Planned Behavior (TPB) (Ajzen, 1985), which developed
out of the Theory of Reasoned Action (Fishbein & Ajzen,
1975). TPB postulates that the immediate precursor to
behavior is intention, defined as the decision or firm
intention to conduct the behavior in the immediate near
future (see Figure 1). According to this theory, intention is,
in turn, influenced by three basic determinants: two of a
personal nature and another reflecting the social influence.
One is the attitude towards the behavior, i.e. the positive
or negative value the person holds towards fulfilling that
action. The second, of a social nature, is the subjective
norm, which is the perception that person has concerning
the social pressures placed upon him or her to perform the
action in question or not. Perceived behavioral control is
the third, the extent that person believes to be in control
over the behavior in question. This component adds greater
power in the capacity to predict (and explain) the behavioral
intention and the very behavior (Ajzen, 1991). In fact, this
antecedent variable, unlike those previous, can directly
predict the behavior depending upon whether this person
is found to be under voluntary control or not and whether
discrepancies exist between the control the person believes
to be in possession of and that control the person actually
has. So, the effect of the perceived behavioral control will
be by the intention when the desired behavior is under
personal control, while the effect will be direct if the
behavioral conduct is not a personal decision. In fact, this
factor is considered an estimation of the actual control one
has over the behavior; however, certain differences can also
exist between the control a person believes to have and the
control he or she actually does. Therefore, the perceived
behavioral control will be a good predictor of the behavior
only if it is comparable with the actual control.
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY
Each of the variables of attitude, subjective norm and
perceived behavioral control, are in turn influenced by other
elements. With respect to the attitude regarding a behavior,
this is a product of the beliefs about the results from such
conduct, and of the assessment the person makes of such
results. The beliefs underlying a person’s attitude towards
a behavior receive the name behavioral beliefs. On the other
hand, the subjective norm derives from a person’s beliefs
that other people or specific groups think that that person
should perform such action or not. These beliefs are called
normative beliefs. But knowing the beliefs a person holds
concerning their important referents is not sufficient for
predicting or understanding their subjective norm. In order
to do this we must also value their motivation to comply
with each of these other people (Ajzen & Fishbein, 1980).
Finally, the perceived behavioral control is in function of
the beliefs about the confidence an individual possesses
regarding their capacities, knowledge, and skills to conduct
a behavior and the beliefs about the ease or difficulty in
carrying out the action. It is based on an evaluation of the
control upon factors that probably will facilitate or inhibit
executing the behavior, multiplied by its occurring frequency.
They include internal control factors (like information,
personal deficiencies, skills, capacities, emotions) and
external control factors (opportunities, dependence upon
others, and obstacles) (Conner & Norman, 1995).
On the other hand, TPB considers such factors like
attitudes towards objectives, personality characteristics,
sociodemographic variables (sex, social class, age), the social
role, status, intelligence, etc., as external values that, although
they could without a doubt influence behavior, they do not
have a direct relationship with it, or they are related with the
intention by the variables included in the theory (Ajzen,
1991). For example, and within the specific context of this
study as is the practice of physical exercise, two factors have
shown to be related as much with the practice of physical
exercise as with the TPB variables, those of self-efficacy and
motivation (Bandura, 1997; Deci & Ryan, 2000). Self-
efficacy, understood as that which I believe capable of doing
with my skills and capabilities in specific situations, is related
to behavioral control, while that of motivation is to attitudes
(Maddux, 2005; Terry & O’Leary, 1995).
TPB has great empirical support showing the relationship
between the variables of this theory and that of physical
exercise (Anderson, Wojcik, Winett, & Williams, 2006; Eng
& Martin-Ginis, 2007; Godin & Kok, 1996). For example,
the recent meta-analysis study by Hagger, Chatzisarantis,
& Biddle (2002) indicates how TPB explained 44.5% of
the variance in physical activity intention and 27.4% of the
variance in physical activity behavior. Ad-hoc questionnaires
were used in the majority of these studies following Azjen’s
recommendations (1985; 1991) of evaluating the different
variables included in TPB, as well as the proposals made
for the “Manual for the construction of questionnaires based
on TPB” (Francis et al., 2004). Such questionnaires presented
803
adequate psychometric properties (Blanchard et al., 2003;
Conner & Sparks, 1996; Jackson, Smith, & Conner, 2009;
Latimer & Martin, 2005). Given that in Spain there are no
validated rating scales integrating the measurement of the
variables included in the TPB into one single instrument,
we feel it opportune to make and validate this instrument
for samples from the Spanish cultural environment. Because
of this, the principal objective in this instrumental study was
to create a tool that evaluates the antecedent variables of
the TPB, i.e. attitude, subjective norm, control and intention,
in getting physical exercise. This study presents the results
from the three independent studies carried out to produce
the Questionnaire of the Theory of Planned Behavior in
Physical Exercise and to gather evidence concerning the
measurements’ validity.
Study 1:
Construction and qualitative evaluation of items
The composition of the items used as a reference the
manual for the construction of the questionnaire based on
TPB by Francis et al. (2004) that supplies information about
designing the questions from each of the variables in the
Theory of Planned Behavior. As such, a data bank of items
was created for the evaluation of the TPB variables in the
practice of physical exercise. The qualitative evaluation of
the items was based on the judgments from a group of experts.
Method
Participants
The panel of experts was formed by seven professionals
in the study of variables involved in TPB or in the field of
questionnaire development, five of whom possessed extensive
knowledge about the different models that are generally
applied to health psychology, and particularly to the TPB
model. The remaining two professionals were experts in the
area of methodology and questionnaire validation.
Procedure
A set of items for each one of the four variables the
model includes was developed based on the previously cited
manual. In all, 21 items were composed, keeping in mind
the semantic definition of the variables under evaluation,
grouped into four subscales: Attitude (eight items),
Subjective !orm (four), Perceived Behavioral Control (five),
and Intention (four). The difference between the number
of items for each subscale is due to the necessity of
attempting to encompass the distinct characteristics of the
semantic definition for all the constructs, without forgetting
about the minimum number of items the manual proposes
804 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN
for each variable either: a minimum number of three items
for the variables subjective norm, perceived behavioral
control and intention, and a minimum of four for attitude.
The manual also suggests the type of response scale for
the variables. As such, the attitude subscale was evaluated
with eight items that included eight pairs of bipolar
adjectives on a response scale from 1 to 7. For the
remaining subscales, the response scale for each one of the
items was of graded response in accordance with quantifiers
from 1 (highly disagree) to 7 (highly agree). The final score
for each subscale was calculated as the average of the scores
obtained for each one of the items in the subscales.
Once the initial bank of items was obtained, it was
presented to the judges for evaluation in two aspects: first,
with respect to the variable belonging to each one of the
items; and second, referring to each item’s composition,
referring to whether they were comprehensible or to point
out aspects that could improve their composition.
Results
In general, the assessments from the group of experts
allowed detecting some deficiencies, principally in the item
composition, and they proposed possible alternatives for
improvement, just like Table 1 shows.
As can be observed in Table 1, there were few item
modifications. The judges indicated that their composition
was quite adequate, although some were modified and some
words eliminated to improve their clarity. One item was
eliminated (If I wanted to, I would easily exercise at least
6 times in the next two weeks.), because it had the same
meaning as another. Lastly, and following the judges
recommendations, one item was added that reflected the
individuals’ motivation to exercise, and it is an important
element for evaluating the subjective norm variable. The
21 remaining items from the study with the group of experts
shaped the set of items empirically proven in Study 2.
Study2:
Item statistical analysis
In this study the authors had the objective of selecting
the final set of the TPB questionnaire items for conducting
physical exercise starting from the analysis of its
psychometric properties.
Method
Participants
One hundred thirty-five people completed the
questionnaire. The average sample age was 27.20 years (SD
= 11.79, range 19-64 years). Out of the total participants,
76 were male (56.3%), 59 were female (43.7%), and 68.9%
were university students.
Instruments
An initial set of 21 items finally constituted the
questionnaire, grouped theoretically into four subscales:
Attitude (eight items), Subjective Norm (four), Behavioral
Control (five) and Intention (four).
Procedure
One hundred-fifty students enrolled in the psychology
and sports sciences degree programs at the Universidad
Miguel Hernández de Elche completed the questionnaire
in a self-administered format. Questionnaires were also
supplied to their parents and grandparents for completion.
A total of 350 questionnaires were administered. After two
weeks, 143 were returned, eight of which were discarded
because they presented elevated numbers of omissions.
Finally, the data from 135 questionnaires (38.57%) were
introduced for the study.
Results
Exploratory factor analysis
In order to learn about the factorial structure of the TPB
questionnaire on physical exercise, exploratory factor
analysis was performed on the 21 selected items by
extraction of principal components with Kaiser
Normalization and Oblimin rotation afterwards, as
relationships between the factors were presumed. The
Kaiser-Guttman method (better known as the Kaiser or K1
rule) was used because it supplies factors that are capable
of explaining, at least, as much variance as an individual
variable. Although simulation studies have shown that its
behavior varies in function of the conditions of application
(Fabrigar, Wegener, MacCallum, & Strahan, 1999; Hakstian
& Muller, 1973; Humphreys, 1964; Jackson, 1993; Linn,
1968; Mote, 1970; Ruiz & San Martín, 1992; Zwick &
Velicer, 1986), the results are not conclusive, this bias
having appeared in situations having few subjects per
variable or when the proportion of variables per factor
increases. On the other hand, although other methods exist,
like the scree plot or Bartlett’s Significance Test, among
others, none has shown to be an unequivocal procedure in
every type of situation. Because of this, the authors conclude
that the K1 rule can be used as a first approximation in the
number of underlying factors, complementing it with
interpretability criteria of the structure, like is our case.
Both the results from Bartlett’s test of sphericity (χ2(210)
= 1925.544, p < .001) as well as the KMO index (.889)
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY 805

Table 1
Modification of the initial set of items
Criteria Original item Modified item

Wording unclear For me, exercising at least 6 times in the next 1.3. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:

Very important to me – No importance Very important – Not important

Insignificant content For me, exercising at least 6 times in the next 1.8. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:

Very intelligent – Very stupid Very absurd - Very intelligent

Item added for a 6. I am motivated to exercise because this is what the

variable majority of people important to me expect.

Eliminated item If I wanted to, I would easily exercise at least

6 times in the next two weeks.

Modified items For me, exercising at least 6 times in the next 1.1. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:

Extremely difficult – Extremely easy Very difficult – Very easy

The majority of the people that are important 2. The majority of the people important to me think I
to me think I should exercise at least 6 times should exercise at least 6 times in the next two
in the next two weeks. weeks.

The majority of the people that are important 5. The majority of the people important to me want
to me want me to exercise at least 6 times in me to exercise at least 6 times in the next two weeks.
the next two weeks.

The majority of the people that are important 12. The majority of the people important to me
to me expect me to exercise at least 6 times in expect me to exercise at least 6 times in the next two
the next two weeks. weeks.

How much personal control do you believe 8. How much control do you believe you have to
you have to exercise at least 6 times in the exercise at least 6 times in the next two weeks?
next two weeks?

I do not believe I have any difficulties to 13. I do not have difficulties to exercise at least 6
exercise at least 6 times in the next two weeks. times in the next two weeks.

I have planned to exercise at least 6 times in 4. I have thought about exercising at least 6 times in
the next two weeks. the next two weeks.

For me, exercising at least 6 times in the next 1.2. For me, exercising at least 6 times in the next
Items maintained two weeks would be: two weeks would be:
from the initial pool Very bad – Very good Very bad – Very good

For me, exercising at least 6 times in the next 1.4. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:
Very unpleasant – Very pleasant Very unpleasant – Very pleasant
806 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN

Table 1 (cont.)
Modification of the initial set of items
Criteria Original item Modified item

For me, exercising at least 6 times in the next 1.5. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:
Very stressful – Very relaxing Very stressful – Very relaxing

For me, exercising at least 6 times in the next 1.6. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:
Very useless – Very useful Very useless – Very useful

For me, exercising at least 6 times in the next 1.7. For me, exercising at least 6 times in the next
two weeks would be: two weeks would be:
Very harmful – Very beneficial Very harmful – Very beneficial

If I wanted to, I could exercise at least 6 3. If I wanted to, I could exercise at least 6 times in
times in the next two weeks. the next two weeks.

It depends entirely upon me whether I exercise 9. It depends entirely upon me whether I exercise at
at least 6 times in the next two weeks. least 6 times in the next two weeks.

I believe I am capable of exercising at least 6 11. I believe I am capable of exercising at least 6

times in the next two weeks. times in the next two weeks.

I will attempt to exercise at least 6 times in 14. I will attempt to exercise at least 6 times in the
the next two weeks. next two weeks.

I will make an effort to exercise at least 6 10. I will make an effort to exercise at least 6 times
times in the next two weeks. in the next two weeks.

I will try to exercise at least 6 times in the 7. I will try to exercise at least 6 times in the next
next two weeks. two weeks.

confirmed the suitability of the correlation matrix to conduct
this analysis. Four components were obtained with
eigenvalues greater than one that explained overall 68.39%
of the variance. The first component offered an eigenvalue
of 8.76 that explained 41.74% of the total variance, grouping
principally the loading from the items that evaluated
Attitude. The second factor presented an eigenvalue of 2.76
explaining 13.14% of the total variance, grouping the items
that evaluated Perceived Behavioral Control. The third
factor attained an eigenvalue of 1.65 that explained 7.86%
of the total variance. The items loaded principally in this
factor corresponded to the Subjective !orm subscale. The
last component offered an eigenvalue of 1.18 and explained
5.64% of the total variance, here primarily grouping the
loading from the items evaluating Intention. Item 1.1. (For
me, exercising at least 6 times in the next two weeks would
be: very difficult —- very easy) did not present a clear
pattern of factor loadings as it was as highly saturated in
the control factor as it was in attitude. Furthermore,
conceptually it could be interpreted as an item as much
about the capacity to exercise as one about the attitude
toward physical exercise, and so it was decided to eliminate
this from the final version of the questionnaire.
Notwithstanding, the reliability of the attitude subscale was
maintained.
Following this elimination, factor analysis for the final
set of 20 items was conducted again, obtaining the factor
loadings per item appearing in Table 2. The factor loadings
adequately grouped together all the items from each one
of the factors measured. As such, the Attitude component
offered an eigenvalue of 8.15, explaining 40.75% of the
total variance and grouped seven items together. Five items
were grouped together in the Perceived Behavioral Control
component, with an eigenvalue of 2.68 that explained
13.40% of the total variance. The Subjective !orm
component presented an eigenvalue of 1.65 and explained
8.25% of the total variance, grouping four items together.
In the last component, Intention, four items were grouped
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY 807

Table 2
Descriptive data for CES-D items in Mexican and Spanish caregivers, with comparison tests and effect size estimates
Components
Items
1 2 3 4

1.2 .679
1.3 .763
1.4 .774
1.5 .744
1.6 .832
1.7 .736
1.8 .819
3 –.723
8 –.661
9 –.647
11 –.677
13 –.797
2 .879
5 .913
6 .591
12 .845
4 –.797
7 –.892
10 –.849
14 –.909

Extraction Method: Principal Components Analysis

Rotation Method: Oblimin
Component 1: Attitude
Component 2: Control
Component 3: Subjective Norms
Component 4: Intention
Note: factorial weights < 0,5 were eliminated

together with an eigenvalue of 1.17 that explained 5.86%
of the total variance.
Therefore, the analyses of the psychometric properties
allowed proposing a final set of 20 items grouped into four
subscales: Attitude, Subjective !orm, Perceived Behavior
Control and Intention, which were adjusted to the theoretical
specifications proposed by TPB.
Analysis of the responses to the items and
estimation of the reliability
Statistical analysis was carried out for the items by
calculating the average, the standard deviation (SD) and
the item-total correlation index (R IT). The internal
consistency index was also calculated by Cronbach’s alpha
for each one of the subscales. Table 3 shows the
psychometric properties for each item within each subscale.
As Table 3 shows, the internal consistency for the
four subscales was very high. The Attitude subscale had
an internal consistency of .878, its items presented average
response values that oscillated between 5.58 for item 1.7,
and 6.8 for items 1.4 and 1.6. All the items presented a
sufficient item-total correlation, superior to .50 in all
cases. The second subscale, Subjective !orm, obtained
an internal consistency coefficient of .852, and the items
having average scores that oscillated between 3.99 for
item 6 and 5.14 for item 2. Moreover, the item-total
correlation values were good, ranging between .54 for
item 6 and item 5 at .77. For the third subscale, Perceived
Behavioral Control, the average response values oscillated
between 5.36 for item 8 and 6.05 for item 11. All the
items presented sufficiently high item-total correlation,
and additionally the internal consistency was adequate
(α = .877). Lastly, the Intention subscale also presented
high reliability (α = .891). The items offered average
response values between 4.94 for item 10 and 5.37 for
items 4 and 14. The items from this subscale also had
high item-total correlations.
808 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN

Table 3
Psychometric properties of the bank ítems of each subscales of TPB questionnarie of physical exercise
Subscale Attitude (α = .88)

Items Mean SD R IT αa

1.2 6.41 1.01 .64 .85

1.3 5.81 1.22 .73 .83
1.4 5.77 1.39 .68 .84
1.5 5.57 1.31 .65 .84
1.6 6.26 .97 .68 .84
1.7 6.45 1.02 .58 .85
1.8 5.85 1.20 .61 .85

Subscale Subjective Norms (α = .85)

2 5.14 1.68 .72 .79

5 4.80 1.71 .77 .77
6 3.99 1.76 .54 .87
12 4.40 1.70 .75 .78

Subscale Control (α = .87)

3 6.01 1.53 .69 ,84

8 5.36 1.50 .66 ,85
9 5.79 1.58 .63 ,96
11 6.05 1.34 .78 ,83
13 5.45 1.86 .76 ,83

Subscale Intention (α = .89)

4 5.37 1.80 .75 .85

7 5.18 1.88 .87 .81
10 4.94 1.85 .54 .93
14 5.37 1.70 .88 .81

SD: Standard Desviation; R IT: correlation item-total;

α a: Cronbach's alpha of scale when ítem is eliminated.

Study 3: women, whose ages ranged from 17-92 years (M = 42.14;

Confirmatory factor analysis and SD = 18.89; Range = 17–92). There were no age differences
external evidence of validity in function of gender (t937 = 0.18; p = .857; δ = .01).

This study provides evidence about the stability of the Instruments

questionnaire’s dimensionality in a sample of the general
population applying confirmatory analysis procedures while
evidence was sought about its validity by means of the
relationship between the scores obtained and other
theoretically related external variables.
Method
Participants
The confirmatory study sample consisted of 944 subjects
from the general population who participated in the
evaluation voluntarily. Of these, 406 were men and 537
For the validity analysis, the subjects were administered
the following instruments along with the refined version
of the TPB questionnaire obtained in the preceding phase.
This is shown in Annex 1.
The General Self-Efficacy Scale (GSE) (Jerusalem &
Schwarzer, 1992), which assesses general perceived self-
efficacy. This work utilized the Spanish adaptation by
Baessler & Schwarzer (1996), which assesses general
competency to efficiently handle a large variety of stressful
situations. This scale obtained good internal consistency,
additionally showing a high positive correlation with self-
esteem and optimism, while it was negative with anxiety,
depression and physical symptoms.
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY
BREQ-2 motivation scale (Behavioral Regulation in
Exercise Questionnaire-2), developed by Markland & Tobin
(2004) and adapted to the Spanish context by Moreno,
Cervelló, & Martínez (2007). This scale consists of 18 items
and 3 factors with moderate reliability coefficients and an
explained total variance of 68.8%. The response scale is
Likert-type of 5 points (1= strongly disagree; 5= strongly
agree). Its three factors are intrinsic regulation, assessed
by four items indicating a subject’s commitment to an
activity due to the pleasure and enjoyment it produces in
the subject; external regulation, comprised of 10 items that
assesses the performance of the individual by an external
incentive; and lastly, amotivation, consisting of four items
and measures the lack of intention for an individual to do
something.
Confirmatory Factor Analysis
In order to confirm the questionnaire’s internal structure,
factor analysis with structural equation models was used.
The estimation of the models was carried out by the AMOS
program, version 15.0. Given that the items deviated
substantially from normality, the majority of their
distributions being asymmetric and platykurtic (K = 223.51),
the Asymptotic Distribution Free (ADF) estimation method
by Browne (1984) was used, whose derivation is not based
on the supposed multivariate normality. In that referring to
atypical cases, these did not occur in the samples.
Student’s t-test for independent samples showed a lack
of differences between men and women in all the
questionnaire’s items, so the analysis was conducted jointly
for both sexes.
In order to apply the analyses and having the intention
of obtaining evidence with distinct samples, the total sample
was randomly divided into two subsamples: Sample 1 (!
= 498; average age = 42.37; SD = 18.72; 43.5% men) and
Sample 2 (! = 446; average age = 41.88; SD = 19.11;
42.6% men). The division was made with the sampling
procedure from SPSS V16.0. No statistically significant
differences were observed between the age from either
sample (t = .396; p = .692; δ = .03).
The test model derived from the findings in the
preceding study so it presented a structure of four factors:
Attitude, comprised of seven items; !orm, formed by four;
Control, with five; and Intention, having four.
The measurement model’s fit was assessed by the χ2
statistic of goodness of fit. Because χ2 tends to be
significant in large samples, the χ2/gl ratio was calculated.
If this ratio is inferior to 6, it is assumed that the model
fit is adequate. This index is called the relative χ2, as it is
calculated to look for the independence from the sample
size. Carmines & McIver (1981) and Kline (1998) establish
a value of 3 for an acceptable model. Furthermore, and
following the recommendation that advises contrasting
various indices to assure the fit of the proposed model,
809
the following goodness of fit indices were kept in mind:
(a) The Goodness of Fit Index, which represents the
percentage of covariances observed that is explained by
the implicit covariances by the model; it must reach a value
equal or superior to .90 (Schumacker & Lomax, 1996);
(b) The Comparative Fit Index, which compares the fit of
the model with a null model supposing that the latent
variables are not related, and requires values equal or super
to .90 (Schumacker & Lomax, 1996); and, (c) the indices
based on the noncentrality parameter, the Root Mean
Square Error of Approximation (RMSEA), which supplies
a measurement of the discrepancy by degrees of freedom.
Values equal to or less than .05 indicate that the model
based on the sample utilized represents the population, and
when values are inferior to .08, the fit is considered
acceptable (Browne & Cudeck, 1992; Jöreskog & Sörbom,
1993).
With regards to the individual fit or the fit of the model
components, factor loadings were used. In this respect, it
must be kept in mind that the sample distribution of the
saturations is not known, which makes assessing the
importance of each variable in relation to the factor difficult,
commonly accepting that there is at least 15% shared
variance by the factor and the variable (Martínez Arias,
1996), which represents saturations of .40.
Initially, the global fit shown by the model was
improvable. The Chi-square test was significant (χ2(116) =
1022, p = .001) and the χ2/gl ratio of 6.2 was sufficiently
superior to that expected. The GFI and CFI adopted values
of .899 and .924, respectively, while the RMSEA estimation
was .074. In the case of subsamples 1 and 2, the values
obtained for the indices were similar to those from the
global sample (χ21 = 611, χ2/gl1 = 3.7; GFI1 = .888, CFI1
= .924, RMSEA1 = .074; χ22 = 605, χ2/gl2 = 3.67; GFI2 =
.878, CFI2 = .919, RMSEA2 = .077).
Due to the inadequate model fit, we explored various
respecification possibilities that were settled in two
modifications conducted in a sequential manner.
– The correlation between the error terms in items 1.3
(Very unpleasant – Very pleasant) and 1.4 (Very stressful –
Very relaxing) was allowed, which is justified because the
modification index reached a very high value and because
it was easily interpretable from the content viewpoint.
– The modification indices also indicated a possible
correlation between the error terms in items 1.1 (Very bad
– Very good) and 1.6 (Very harmful – Very beneficial).
Upon assessing the content of these items it was concluded
they were redundant, and so item 1.1 was eliminated since
it was that of the two with the lowest loading, as well as
being the item involving the highest values in terms of the
standardized residuals.
A new CFA was performed with the respecified 19-item
model, whose results show a better fit with the data, in both
the global sample (χ2 = 821, χ2/gl = 5.6; GFI = .914, CFI
= .936, RMSEA = .067), as well as in the subsamples (χ21
810 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN

= 472, χ2/gl1 = 3.2, GFI1 = .909, CFI1 = .941, RMSEA1 =

.067; χ22 = 504, χ2/gl2 = 3.5, GFI2 = .891, CFI2 = .930,
RMSEA2 = .074). The graphic representation of this model
is in Figure 2.

Internal consistency

The calculation of the internal consistency of the

different dimensions obtained was performed utilizing
Omega coefficient (Ω) by Heise & Bohrnstedt (1970). Its
calculation is based on the results of factor analysis and
given by:

n – ∑ hi2
Ω = –––––––––––
n – 2 ∑ rij

The results obtained reflect high consistency, as much

in the total sample as in samples 1 and 2. A value of .88
was reached in these three cases for the Attitude factor,
while for !orm they were .85, .86 and .84; for Control,
they were .79, .81 and .76; and lastly, they were .89 in the
three samples for the Intention factor (See Figure 2). The
internal consistency values were high for all the test’s
dimensions.

Correlation of factors and factor loadings of the

items

Given the specified relationships in the model, high

correlations were expected between the four factors. Just
as Figure 2 shows, high correlations are observed between
the factors of Attitude, Intention and Control, whereas they
are low in the relationships of all of them with the !orm
factor. This result repeats itself as much in subsample 1
(rAI1 = .63; rAC1 = .58; rIC1 = .78; rA!1 = .27; rI!1 = .34;
rC!1 = .24) as in subsample 2 (rAI1 = .55; rAC1 = .54; rIC1
= .77; rA!1 = .22; rI!1 = .31; rC!1 = .18)
With respect to the factor loadings, all are high,
oscillating between .54 and .94 for the global sample, while
between .55 and .94 for subsample 1. They are between
.50 and .94 for subsample 2.
Evidence of Validity
This section provides evidence concerning the
relationships between the measurements contributed by the
written questionnaire and other external variables
theoretically related to the constructs included in the TPB.
In order to conduct these analyses we used Pearson’s
correlation coefficient. To avoid bias in the sample sizes
and as a measurement of the effect size, we followed the
suggestions by Cohen (1988), who classifies the effect size
as being small if r ≈ .10, medium if r ≈ .30, and large if
r ≈ .50.
Figure 2. Model of measure of global sample.
The results appearing in Table 4 show relationship patterns
coherent with the theory. One the one hand, Self-efficacy
presented the greatest relationship with the Control dimension,
while the lowest occurred in the !orm factor. On the other
hand, the motivation variables also presented the expected
results. In this way, intrinsic motivation, which refers to the
search for pleasure and satisfaction in sporting activity
participation (Deci & Ryan, 1985), was related with all the
TPB dimensions except for that of !orm, which is the model’s
external reference, a pattern that is repeated in the case of
identified motivation, i.e., the interest the subject displays in
sporting activity participation in order to attain goals considered
relevant for the subject’s personal development (Deci & Ryan,
1985). For its part, external motivation, which is the interest
in participating in sporting activities for awards or compensation
(Deci, 1975), presented positive relations with !orm; for the
remaining dimensions they were negative. Lastly, amotivation,
understood as the lack of motivation or intention to continue
exercising, did not show a relationship with !orm, while it
did do so with the other factors with negative signs.
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY 811

Table 4
Correlations between dimensions and criteria variables
Self-efficacy Intrinsic Motivation Identify Motivation External Motivation Amotivation

Global Attitude .21** .44** .41** –.21** –.34**

Norm .05 .03 .11** .29** .04
Control .28** .39** .38** –.31** –.31**
Intention .16** .30** .42** –.10** –.30**
Sample 1 Attitude .16** .49** .51** –.24** –.43**
Norm .05 .06 .12** .28** .04
Control .20** .34** .38** –.32** –.37**
Intention .11** .29** .40** –.17** –.31**
Sample 2 Attitude .26** .40** .31** –.18** –.25**
Norm .05 .01 .10 .30** .05
Control .37** .46** .37** –.31** –.24**
Intention .22** .30** .44** –.03 –.29**

** p < .01

Discussion Omega coefficient (Ω). Furthermore, these results are similar

In this work the authors’ objectives were to develop
an instrument to assess the different variables included in
the TPB (attitude, subjective norm, control and intention)
within the specific context of practicing physical exercise.
This objective was covered with the implementation of
the three studies presented that include the construction
and qualitative assessment of the items (study 1), the
analysis of the measurements’ psychometric properties,
along with the estimation of their reliability (study 2), and
the study of the instrument’s dimensionality, as well as
obtaining external evidence of its validity (study 3). On
one hand, the results from the different analyses carried
out support the factorial structure of four factors proposed
by the TPB model (Ajzen, 1985), showing all the goodness
of fit index values within the ranges established as being
adequate. Notwithstanding, in order to corroborate the
good fit of the estimated model, the individual parameters
were also assessed, given the existing controversy around
the choice of cut-off points for the goodness of fit indices
in a conventional or “routine” manner (Hu & Bentler,
1999). As it is known, these are highly contested due to
a lack of empirical evidence and a logical convincing base
for its acceptance (Yu, 2002), in addition to the variability
in its behavior under distinct conditions, like the sample
size, its distributional properties or the estimation method
utilized (Bentler, 1990; Bentler & Yuan, 1999; Bollen,
1990; Curran, West, & Finch, 1996; Fan, Thompson, &
Wang, 1999; Hu, Bentler, & Kano, 1992; Nevitt &
Hancock, 2000; Sugawara & MacCallum, 1993; Yuan &
Bentler, 1998).
On the other hand, the four subscales making up this
instrument showed adequate internal consistency as much
in the initial phase, estimated with the coefficient of
Cronbach’s alpha, as in the confirmatory study with the
to those found in other work using an ad-hoc questionnaire
(Chatzisarantis & Hagger, 2008; Jackson et al., 2009). The
adequate psychometric properties that the items forming
this questionnaire were shown to have can be due to the
meticulous work done in the questionnaire’s preparation.
This instrument, just like previously described, was
accomplished in a systematic manner following the criteria
proposed by the manual “Constructing Questionnaires Based
on The Theory of Planned Behavior” (Francis et al., 2004),
which in turn guarantees the validity of the instrument’s
content.
Lastly, the analysis of the relationships with other
variables showed evidence of convergent validity in the
correlations of self-efficacy and motivation with the different
subscales making up the TPB-based questionnaire, just like
as was expected according to previous studies (Bandura,
1997; Neupert, Lachman, & Whitbourne, 2009; Rhodes,
Blanchard, Matheson & Coble, 2006). Self-efficacy was
related in a positive manner with control, attitude and
intention, whereas amotivation was related in a negative
form with attitude, control and intention. For its part,
external motivation showed positive relationships with
subjective norm.
To conclude, and although further research is still
required, with samples from different sources to analyze
in their full extent the psychometric guarantees of this test,
we have an instrument coming to occupy an existing
knowledge gap of great relevance today in the field
researching the practice of physical exercise with the TPB.
An instrument has been constructed that meets the objectives
for which it was developed, providing research professionals
from different health promotion fields a tool with which to
acquire reliable and valid measurements. As a limitation
of the study and following phase in this research, the
assessment of the predictive validity of this scale should
812 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN
be conducted specifically from its capacity to explain the
practice of physical exercise. For this, some caution is
recommended in the use of this instrument in the scope of
the assessment of physical exercise as the scale’s validation
is still incomplete and unfinished.
References
Ajzen, I. (1985). From intentions to actions: A theory of planned
behavior. In J. Kuhi & J. Beckmann (Eds.), Action-control:
From cognition to behavior (pp.11-39). Heidelberg, Germany:
Springer.
Ajzen, I. (1991). The theory of planned behavior. Organizational
Behavior and Human Decision Processes, 50, 179–211.
http://dx.doi.org/10.1016%2F0749-5978%2891%2990020-T
Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and
predicting social behavior. Nueva York, NY: Prentice Hall.
Anderson, E. S., Wojcik, J. R., Winett, R. A., & Williams, D. M.
(2006). Social-cognitive determinants of physical activity:
The influence of social support, self-efficacy, outcome
expectations, and self-regulation among participants in a
church-based health promotion study. Health Psychology, 25,
510–520. http://dx.doi.org/10.1037%2F0278-6133.25.4.510
Baessler, J., & Schwarzer, R. (1996). Evaluación de la autoeficacia:
Adaptación española de la escala de autoeficacia general
[Measuring generalized self-beliefs: A Spanish adaptation of
the General Self-Efficacy scale]. Ansiedad y Estrés, 2, 1–8.
Bandura, A. (1997). Self-efficay: Toward a unifying theory of
behavioral change. Psychological review, 84, 191–215.
http://dx.doi.org/10.1037%2F%2F0033-295X.84.2.191
Bentler, P. M. (1990). Comparative fit indexes in structural models.
Psychological Bulletin, 107, 238–246. http://dx.doi.org/10.1037
%2F%2F0033-2909.107.2.238
Bentler, P. M., & Yuan, K. H. (1999). Structural equation modeling
with small samples: Test statistics. Multivariate Behavioral
Research, 34, 181–197. http://dx.doi.org/10.1207%2FS15327
906Mb340203
Biddle, S. J. H., Fox, K. R., & Boutcher, S. H. (2000). Physical
activity and psychological wellbeing. Londres, England:
Routledge.
Blair, S. N., & Brodney, S. (1999). Effects of physicial inactivity
and obesity on morbidity and mortality: Current evidence and
research issues. Medicine and Science in Sports and Exercise,
31, 646–662. http://dx.doi.org/10.1097%2F00005768-199911001-
00025
Blanchard, C. M., Courneya, K. S., Rodgers, W. M., Fraser, S.
N., Murray, T. C., & Daub, B. (2003). Is the theory of planned
behavior a useful framework for understanding exercise
adherente during phase II cardiac rehabilitation? Journal of
Cardiopulmonary Rehabilitation, 23, 29–39. http://dx.doi.org/
10.1097/00008483-200301000-00007
Bollen, K. A. (1990). Overall fit in covariance structure models:
Two types of sample size effects. Psychological Bulletin, 107,
256–259. http://dx.doi.org/10.1037%2F%2F0033-2909.107.2.256
Browne, M. V. (1984). Asymptotically distribution-free methods
for the analysis of covariance structures. British Journal of
Mathematical and Statistical Psychology, 37, 62–83.
http://dx.doi.org/10.1111%2Fj.2044-8317.1984.tb00789.x
Browne, M. V., & Cudeck, R. (1992). Alternative ways of assessing
model fit. Sociological Methods and Research, 21, 230–258.
http://dx.doi.org/10.1177%2F0049124192021002005
Carmines, E. G., & McIver, J. P. (1981). Analyzing models with
unobserved variables: Analysis of covariance structures. In
G. W. Bohmstedt & E. F. Borgatta (Eds.), Social Measurement
(pp. 65-115). Thousand Oaks, CA: Sage Publications.
Chatzisarantis, N., & Hagger, M. (2008). Influences of personality
traits and continuation intentions on physical activity participation
within the theory of planned behavior. Psychology & Health,
23, 347–367. http://dx.doi.org/10.1080%2F14768320601185866
Cohen, J. (1988). Statistical power analysis for the behavioral
science. Hillsdale, NJ: Lawrence Erlbaum Associates.
Conner, M., & Norman, P. (1995). The role of social cognition
in health behaviors. In M. Conner & P. Norman (Eds.),
Predicting Health Behavior: Research and practice with social
cognition models (pp. 1-22). Buckingham, England: Open
University Press.
Conner, M., & Sparks, P. (1996). The theory of planned behavior
and health behaviors. In M. Conner & P. Norman (Eds.),
Predicting Health Behavior: Research and practice with social
cognition models (pp. 121-162). Buckingham, England: Open
University Press.
Curran, P. S., West, S. G., & Finch, J. F. (1996). The robustness
of test statistics to nonnormality and specification error in
confirmatory factor analysis. Psychological Methods, 1, 16–
29. http://dx.doi.org/10.1037%2F%2F1082-989X.1.1.16
Deci, E. L. (1975). Intrinsic motivation. New York, NY: Plenum
Press.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-
determination in human behavior. New York, NY: Plenum Press.
Deci. E. L., & Ryan, R. M. (2000). The “what” and “why” of
goal pursuits: Human needs and the self-determination of
behavior, Psychological Inquiry, 11, 227–268. http://dx.doi.org/
10.1207%2FS15327965PLI1104_01
Eng, J. J., & Martin-Ginis, K. A. (2007). Using the theory of
planned behavior to predict leisure time physical activity
among people with chronic kidney disease. Rehabilitation
Psychology, 52, 435–442. http://dx.doi.org/10.1037%2F0090-
5550.52.4.435
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan,
E. J. (1999). Evaluating the use of exploratory factor analysis
in psychological research. Psychological Methods, 4, 272–
299. http://dx.doi.org/10.1037%2F%2F1082-989X.4.3.272
Fan, X., Thompson, B., & Wang, L. (1999). Effects of sample
size, estimation methods, and model specification on structural
equation modeling fit indexes. Structural Equation Modeling,
6, 56–83. http://dx.doi.org/10.1080%2F10705519909540119
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and
behavior: An introduction to theory and research. Nueva York,
NY: Addison-Wesley.
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY
Francis, J. J., Eccles, M. P., Johnston, M., Walker, A., Grimshaw,
J., Foy, R., … Bonetti, D. (2004). Constructing questionnaires
based on the theory of planned behavior: A manual for health
services researchers. Newcastle, England: Centre for Health
Services Research.
Godin, G., & Kok, G. (1996). The theory of planned behavior: A
review of its applications to health-related behaviors. American
Journal of Health Promotion, 11, 87–98. http://dx.doi.org/
10.4278%2F0890-1171-11.2.87
Hagger, M. S., Chatzisarantis, N. L. D., & Biddle, S. J. H. (2002).
A meta-analytic review of the theories of reasoned action and
planned behavior in physical activity: Predictive validity and
the contribution to additional variables. Journal of Sport and
Exercise Psychology, 24, 3–32.
Hakstian, A. R., & Muller, V. J. (1973). Some notes on the number
of factors problem. Multivariate Behavioral Research, 461–
475. http://dx.doi.org/10.1207%2Fs15327906mbr0804_4
Heise, D. R., & Bohrnstedt, G. W. (1970). Validity, invalidity and
reliability. In E. F. Borgatta & G. W. Bohrnstedt (Eds.),
Sociological Methodology (pp. 104-129). San Francisco, CA:
Jossey-Bass. http://dx.doi.org/10.2307%2F270785
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in
covariance structure analysis: Conventional criteria versus
new alternatives. Structural Equation Modeling, 6, 1–55.
http://dx.doi.org/10.1080%2F10705519909540118
Hu, L., Bentler, P. M., & Kano, Y. (1992). Can test statistics in
covariance structure analysis be trusted? Psychological Bulletin,
112, 351–362. http://dx.doi.org/10.1037%2F%2F0033-2909.
112.2.351
Humphreys, L. G. (1964). Number of cases and number of factors:
An example where N is very large. Educational and
Psychological Measurement, 24, 457–466. http://dx.doi.org/
10.1177%2F001316446402400302
Jackson, D. A. (1993). Stopping rules in principal components
analysis: A comparison of heuristical and statistical approaches.
Ecology, 74, 2204–2214. http://dx.doi.org/10.2307%2F1939574
Jackson, C., Smith, A., & Conner, M. (2009). Applying an
extended version of the theory of planned behavior to physical
activity. Journal of Sports Sciences, 21, 119–133.
http://dx.doi.org/10.1080%2F0264041031000070976
Jerusalem, M., & Schwarzer, R. (1992). Self-efficacy as a resource
factor in stress appraisal processes. In R. Schawarzer (Ed.),
Self-efficay: Thought control of action. Washington, DC:
Hemisphere.
Jiménez, M., Martínez, P., Miró, E., & Sánchez, A. I. (2008).
Bienestar psicológico y hábitos saludables: ¿están asociados
con la práctica de ejercicio físico? [Psychological well-being
and healthy habits: Are they associated with physical exercise
practice?]. International Journal of Clinical and Health
Psychology, 8, 85–102.
Jöreskog, K., & Sörbom, D. (1993). LISREL 8: Structural equation
modeling with the SIMPLIS command language. Hillsdale,
NJ: Erlbaum.
Kline, R. B. (1998). Principles and practice of structural equation
modeling. New York, NY: Guilford Press.
813
Latimer, A. E., & Martin, A. (2005). The theory of planned
behavior in prediction of leisure time physical activity among
individuals with spinal cord injury. Rehabilitation Psychology,
50, 389–396. http://dx.doi.org/10.1037%2F0090-5550.50.4.389
Lawlor, D. A., & Hopker, S. W. (2001). The effectiveness of
exercise as an intervention in the management of depression:
Systematic review and meta-regression analysis of randomised
controlled trials. British Medical Journal, 322, 763–767.
http://dx.doi.org/10.1136%2Fbmj.322.7289.763
Linn, R. L. (1968). A Monte Carlo approach to the number of
factors problem. Psychometrika, 33, 37–71. http://dx.doi.org/
10.1007%2FBF02289675
Maddux, J. E. (2005). Self-efficacy. In C. R. Snyder & S. J. Lopez
(Eds.), Handbook of Positive Psychology (pp. 277-287).
Oxford: Oxford University Press.
Marcus, B. H., Dubbert, P. M., Forsyth, L. H., McKenzie, T. L.,
Stone, E. J., Dunn, A. L., & Blair, S. N. (2000). Physical activity
behavior change: Issues in adoption and maintenance. Health
Psychology, 19, 32–41. http://dx.doi.org/10.1037%2F%2F0278-
6133.19.Suppl1.32
Markland, D., & Tobin, V. (2004). A modification to the behavioral
regulation exercise questionnaire to include and assessment
of amotivation. Journal of Sport & Exercise Psychology, 26,
191–196.
Martínez Arias, M. R. (1996). Psicometría: Teoría de los tests
psicológicos y educativos [Psicometry: Theory of psychological
and educational tests]. Madrid, Spain: Síntesis.
McAuley, E., Mihalko, S. L., & Bane, S. M. (1997). Exercise and
self-esteem in middle-aged adults: Multidimensional
relationships and physical fitness and self-efficacy influences.
Journal of Behavioral Medicine, 20, 67–83. http://dx.doi.org/
10.1023/A:1025591214100
Moreno, J. A., Cervelló, E. M., & Martínez, A. (2007). Measuring
self-determination motivation in a physical fitness setting:
Validation of the Behavioral Regulation in Exercise
Questionnaire-2 (BREQ-2) in a Spanish sample. The Journal
of Sport Medicine and Physical Fitness, 47, 366–378.
Mote, T. A. (1970). An artifact of the rotation of too few factors:
Study orientation vs. trait anxiety. Revista Internacional de
Psicología, 37, 61–91.
Netz, Y., Wu, M. J., Becker, B. J., & Tenebaum, G. (2005). Physical
activity and psychological well-being in advanced age: A meta-
analysis of intervention studies. Psychology and Aging, 20,
272–284. http://dx.doi.org/10.1037%2F0882-7974.20.2.272
Neupert, S. D., Lachman, M. E., & Whitbourne, S. B. (2009).
Exercise self-efficacy and control beliefs: Effects on exercise
behavior after an exercise intervention for older adults. Journal
of Aging and Physical Activity, 17, 1–16.
Nevitt, J., & Hancock, G. R. (2000). Improving the root mean
square error of approximation for nonnormal conditions in
structural equation modeling. The Journal of Experimental
Education, 68, 251–268. http://dx.doi.org/10.1080%2F002209
70009600095
Rhodes, R., Blanchard, C., Matheson, D., & Coble, J. (2006).
Disentangling motivation, intention and planning in the
814 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN
physical activity domain. Psychology of Sport and Exercise,
7, 15–27. http://dx.doi.org/10.1016%2Fj.psychsport.2005.08.011
Ruiz, M. A., & San Martín, R. (1992). Una simulación sobre el
comportamiento de la regla K1 en la estimación del número
de factores [A simulation about the behaviour of K1 rule in
the estimation of factor numbers]. Psicothema, 4, 543–550.
Shumacker, R. E., & Lomax, R. G. (1996). A beginner’s guide to
structural equation modeling. Mahwah, NJ: Erlbaum.
Sjöström, M., Oja, P., Hagströmer, M., Smith, B. J., & Bauman, A.
(2006). Health-enhancing physical activity across European Union
countries: The Eurobarometer study. Journal of Public Health,
14, 291–300. http://dx.doi.org/10.1007%2Fs10389-006-0031-y
Sniehotta, F. F., Scholz, U., & Schwarzer, R. (2005). Brindging
the intention-behavior gap: Planning, self-efficacy, and action
control in the adoption and maintenance of physical exercise.
Psychology and Health, 20, 143–160. http://dx.doi.org/10.1080/
08870440512331317670
Sugawara, H. M., & MacCallum, R. C. (1993). Effect of estimation
method on incremental fit indexes for covariance structure
models. Applied Psychological Measurement, 17, 365–377.
http://dx.doi.org/10.1177%2F014662169301700405
Taras, H. (2005). Physical activity and student performance at
school. The Journal of School Health, 75, 214–218.
http://dx.doi.org/10.1111%2Fj.1746-1561.2005.00026.x
Terry, D. J., & O’Leary, J. E. (1995). The theory of planned
behavior: The effects of perceived behavioral control and self-
efficacy. British Journal of Social Psychology, 34, 199–220.
http://dx.doi.org/10.1111%2Fj.2044-8309.1995.tb01058.x
WHO (2002). The world heatlh report 2002-Reducing risks,
promoting healthy life. Geneva, Switzerland: Author.
WHO (2004). World strategy about daily diet, physical activity
and heatlh. Geneva, Switzerland: Author.
WHO (2006). Physical activity and health in Europe: evidence for
action. Copenhagen, Denmark: WHO Regional Office for Europe.
Yu, C. Y. (2002). Evaluating cutoff criteria of model fit indices
for latent variable models with binary and continuous
outcomes. (Doctoral dissertation). University of California,
Los Angeles, CA. Retrieved from http://statmodel.com/
download/Yudissertation.pdf.
Yuan, K.-H., & Bentler, P. M. (1998). Normal theory based test
statistics in structural equation modelling. British Journal of
Mathematical and Statistical Psychology, 51, 289–309.
http://dx.doi.org/10.1111%2Fj.2044-8317.1998.tb00682.x
Zwick, W. R., & Velicer W. F. (1986). A comparison of five rules
for determining the number of components to retain.
Psychological Bulletin, 99, 432–442. http://dx.doi.org/
10.1037%2F%2F0033-2909.99.3.432
Received June 16, 2010
Revision received May 30, 2011
Accepted June 13, 2011
 VALIDATION OF A QUESTIONNAIRE IN PHYSICAL ACTIVITY 815

APPE DIX

Questionnaire of the Theory of Planned Behavior regarding the practice of physical exercise.
816 TIRADO, NEIPP, QUILES, AND RODRÍGUEZ-MARÍN