Overcoming Procrastination:

The Effect of Implementation Intentions1

Shane Gregory Owens,2 Christine G. Bowman, and Charles A. Dill

Hofstra University

A hypothesized solution for procrastination is the formation of an implementation

intention (Van Eerde, 2000). University students (N = 152) were assessed using the
Aitken Procrastination Inventory (Aitken, 1982) and were asked to report to an
experiment. Half of the participants formed implementation intentions to attend. An
odds ratio indicated that participants who formed implementation intentions were
nearly 8 times more likely to keep their appointments than those who did not. Low
procrastinators reported more often for the experiment than did high procrastina-
tors (Low = 49.4%; High = 30.1%), c2(1, N = 152) = 5.84, p < .016. The interaction
between implementation intentions and procrastination was not significant,
c2(1, N = 152) = 0.28, p < .60.

Procrastination has been defined as the purposeful postponement or

delaying of the performance of a task or the making of a decision (Ferrari,
2001; Milgram, Mey-Tal, & Levison, 1998). It has also been described as a
self-regulatory style that delays the start or completion of a task (Ferrari &
Tice, 2000). Van Eerde (2000) defined procrastination “as the avoidance of
the implementation of an intention”(p. 374).
Procrastinators may be a form of inclined abstainers (Sheeran, 2002), who
intend to act but do not follow through with their intentions. Schouwenburg
(2004) highlighted the difference between state procrastination, in which
avoidance behavior is task-specific; and trait procrastination, as “the ten-
dency to postpone that which is necessary to reach some goal” (p. 5). Pos-
sessing this trait increases the probability that one will delay action initiation
in a variety of instances. It is clear from all of these definitions that the
essential aspect of procrastination is not performing a behavior when an
adequate opportunity presents itself.
The negative consequences of procrastination include internal conse-
quences, such as depression, anxiety, and guilt (Burka & Yuen, 1983; Flett,

1
Shane Gregory Owens is associate director of Psychological Services at Farmingdale State
College and an assistant professor at St. John’s University. Christine Bowman is now a graduate
student at Fordham University. This paper is based on the second author’s honors thesis
research, which was supervised by the third author.
2
Correspondence concerning this article should be addressed to Shane Gregory Owens, 283
Commack Road, Suite LL 2, Commack, NY 11725. E-mail: [email protected]

366

Journal of Applied Social Psychology, 2008, 38, 2, pp. 366–384.

© 2008 Copyright the Authors
Journal compilation © 2008 Blackwell Publishing, Inc.
 OVERCOMING PROCRASTINATION 367

Blankstein, & Martin, 1995); and external consequences, such as lateness at

work (Van Eerde, 2000), fines for late library books, failed marriages (Burka
& Yuen, 1983), and poor performance on class exercises, final exams, and
lower overall course grades (Steel, Brothen, & Wambach, 2001). Procrasti-
nation is also associated with negative effects on physical health (Sirois,
Melia-Gordon, & Pychyl, 2003; Tice & Baumeister, 1997). On a larger scale,
procrastination has negative effects on political and economic decisions
(Holland, 2001; Kegley, 1989).
According to a meta-analysis by Van Eerde (2003), there is a positive,
direct relationship between procrastination and missing a deadline, and a
negative, direct relationship between procrastination and task preparation.
In addition, there are moderated, moderately negative relationships between
procrastination and course grade, and procrastination and overall grade
point average (Van Eerde, 2003). In terms of affective variables, Van Eerde
found a consistent positive relationship between procrastination and depres-
sion and between procrastination and anxiety.3
Procrastination is a prevalent problem, especially among young adults. It
is estimated that anywhere between 50% and 95% of college students engage
in procrastination (Day, Mensink, & O’Sullivan, 2000; Ellis & Knaus, 1977;
Knaus, 2000; Potts, 1987; Pychyl, Lee, Thibodeau, & Blunt, 2000; Solomon
& Rothblum, 1984). Harriott and Ferrari (1996) reported that approximately
15% to 20% of adults engage in chronic procrastination.
Though the amount of literature regarding procrastination has increased
over the past decade, most of this literature is devoted to finding the causes or
predictors of procrastination, and not to finding strategies to overcome it.
Those sources that do focus on preventing procrastination are mostly limited
to self-help literature. Some authors have outlined techniques for overcoming
procrastination, but have offered little empirical support for any of these
techniques. Examples include Ferrari (2001), who outlined 10 strategies,
including learning organizational skills, using reminder notes, and breaking
complex goals into smaller, more manageable pieces; and Van Eerde (2000)
who advocated the training of self-regulation skills, altering the context in
which tasks are supposed to be performed, and planning. Schouwenburg,
Lay, Pychyl, and Ferrari (2004) described techniques for assisting academic

3
Some researchers have identified a type of procrastinator who is engaging in a behavior that
enhances his or her functioning. Chu and Choi (2005) differentiated between passive procrasti-
nators and active procrastinators. As opposed to the traditional negative procrastinator, an
active procrastinator makes decisions to delay action and is likely to achieve more satisfactory
outcomes than is the passive procrastinator. In addition, Schouwenburg (2004) stated that some
dilatory behavior may, in some situations, be wise. For the purposes of the current experiment,
no distinction was made between these two types of procrastinators, as the experimenters were
interested only in procrastination behavior that carries some measure of harm.
368 OWENS ET AL.

procrastinators to overcome their difficulties. Their work utilized research on

procrastination and application of therapeutic techniques (e.g., cognitive-
behavior therapy) to outline methods for those who counsel procrastinators
in academic settings.
The current study examines the relationship between intentions and
behavior, which has been established as an important issue in procrastina-
tion. In addition, it seeks to apply Gollwitzer’s (1999) concept of implemen-
tation intentions to the effort to overcome procrastination and its deleterious
effects.

Decision and Action

Gollwitzer (1999) delineated two phases involved in the decision to act

and the performance of actions. The first phase is motivational and involves
the selection of an action to bring about certain personal and social conse-
quences. The second phase is volitional and involves the performance of
selected actions. As it has been established that behavior involves the moti-
vational aspect of intending to behave in a certain way and the volitional
aspect of acting on those intentions, it is important to use an understanding
of both phases in any strategy to overcome procrastination.
While Ajzen’s (1985, 1991; Ajzen & Madden, 1986) theory of planned
behavior (TPB) provides insight into the motivational phase of behavioral
enactment, Gollwitzer’s (1999) theory of implementation intentions provides
a strategy to overcome difficulties during the volitional phase. Procrastina-
tion seems to be a phenomenon that involves both phases of behavioral
enactment, as behavioral enactment is the result of intentions to behave
(motivational) and actions toward goal achievement (volitional). A full
understanding of procrastination would, therefore, have to involve both
motivational and volitional factors.

Behavioral Intentions

Ajzen’s (1985, 1991; Ajzen & Madden, 1986) TPB states that the intention
to perform a behavior is an important proximal determinant for that behav-
ior. Sirois (2004) suggested that procrastinators do not engage in healthy
behaviors because they have weaker intentions to perform such behaviors.
Van Eerde’s (2000) definition of procrastination indicates that it is a break-
down of the implementation of an intention to act, and Steel et al. (2001)
stated that procrastination reflects a discrepancy between intentions and
actions, and that procrastinators show the largest intention–action gap.
 OVERCOMING PROCRASTINATION 369

All of these explanations for procrastination highlight the importance of

behavioral intentions as a predictor of behavioral enactment. Reviews of
research on the TPB (Ajzen, 1985, 1991) indicate that, while the theory has
good predictive validity, it often accounts for only 20% to 30% of behavioral
enactment. Sheeran, Webb, and Gollwitzer (2005) highlighted the impor-
tance of strong goal intentions in completing goal-related tasks. Taken
together, this evidence suggests that, while good intentions may be important
in explaining why a person behaves, they may not be sufficient conditions for
behavioral enactment, and procrastinators may have problems in translating
even strong intentions into action.
While the TPB accounts for intention to perform a certain action, behav-
ioral intentions are not sufficient in and of themselves to explain behavioral
enactment. Implementation intentions have been shown to augment the
predictive validity of the TPB (e.g., Orbell, Hodgkins, & Sheeran, 1997).

Implementation Intentions

Gollwitzer (1999) developed implementation intentions in order to effec-

tively translate intention into action. Implementation intentions augment goal
intentions by adding specifications of when and where to perform a specified
behavior. For example, the goal intention “I will write the first draft of my
article” becomes “I will write the first draft of my article in my office at 6:00
p.m. on Sunday” when supplemented with an implementation intention.
Gollwitzer and Sheeran (in press) reviewed problems in goal pursuit
related to procrastination. In terms of starting a task, they mentioned prob-
lems with remembering to act, utilizing appropriate opportunities for action,
and overcoming reluctance to act. Gollwitzer (1999) indicated that forming
an implementation intention causes the memory for the context specified in it
to become highly activated and more readily accessible, making it the kind of
plan that is likely to overcome distraction, which, according to Van Eerde
(2000), is important in overcoming procrastination. In addition, the context
specified by the “when” and the “where” of the implementation intention is
assumed to elicit the behavior specified in the implementation intention
“immediately, efficiently (i.e., not requiring much processing capacity), and
without conscious intent” (Brandstätter, Lengfelder, & Gollwitzer, 2001,
p. 947). In short, by passing control of the behavior to environmental cues
contained in the implementation intention, the behaviors demand few cog-
nitive resources and the effects of distractions, competing and other difficult
goals, other habits, and procrastination are minimized (Brandstätter et al.,
2001; Gollwitzer & Brandstätter, 1997). In addition, Van Eerde (2000) men-
tioned implementation intentions as a way to overcome procrastination by
370 OWENS ET AL.

visualizing an appropriate time and place for action ahead of time, and using
this context efficiently when it arrives.
There has been much research demonstrating the efficacy of implementa-
tion intentions in initiating behaviors. For example, Verplanken and Faes
(1999) found that implementation intentions helped participants to initiate a
healthy diet. Sheeran and Orbell (1999) found that implementation intentions
enhanced the taking of vitamins, helped patients return to regular physical
activity quickly following surgery (Orbell & Sheeran, 2000), and ensured that
women would perform breast self-examination (Orbell et al., 1997). In addi-
tion, Dill, Owens, and colleagues (Dill, Owens, Homburger, Travers, &
Lancaster, 2002; Owens, 2003; Owens, Dill, Levine, & Goldfarb, 2001, 2002)
found that forming implementation intentions helped college students to
complete academic-related tasks, such as sending weekly e-mails to instruc-
tors and keeping a weekly journal. Furthermore, Koole and van’t Spijker
(2000) found that participants who formed implementation intentions wrote
a report of experiences during a specific day significantly faster than those
participants who did not form implementation intentions.4

Rationale and Hypotheses for the Current Study

Many researchers (Sirois, 2004; Steel et al., 2001; Van Eerde, 2000) have
indicated that behavioral intentions play a role in procrastination. While
some (e.g., Van Eerde, 2000) have indicated that implementation intentions
may have some utility in counteracting procrastination, there have been no
published studies to date examining their effectiveness in counteracting pro-
crastination. The current study is undertaken to examine the relationship
between behavioral intentions and behavioral enactment in procrastinators
and to examine the effects of forming implementation intentions on keeping
an appointment in college students who identify themselves as high or low
procrastinators. We hypothesize the following:
Hypothesis 1. Individuals who form implementation intentions
will perform the behavior more frequently than those who do
not.
Hypothesis 2. Individuals who rate themselves as low procras-
tinators will perform the behavior more frequently than those
who rate themselves as high procrastinators.
4
See Gollwitzer and Sheeran (in press) for a meta-analytic review in which they calculated
the implementation intention effect across 94 studies with 8,461 participants (d = .65), and
Koestner, Lekes, Powers, and Chicoine (2002) for a meta-analytic review in which they calcu-
lated the implementation intention effect across 13 published studies (d+ = .54).
 OVERCOMING PROCRASTINATION 371

Hypothesis 3. Individuals with high intentions will be more

likely to enact the behavior than those with low intentions.
Hypothesis 4. The best fitting model of behavioral enactment
will contain main effects for procrastination, behavioral inten-
tions, and implementation intentions.
Hypothesis 5. Individuals who have strong intentions and who
form implementation intentions will have the highest rate of
behavioral enactment.

Method

Participants

The participants were 152 university students (96 female, 56 male) who
volunteered to participate for extra course credit. The students were pri-
marily from the upper divisions (60.4% were juniors or seniors), most of them
were psychology majors (63.2%), and their mean grade point average (GPA)
was 3.07 (SD = 0.48). The students were treated in accordance with the
ethical standards of the APA.5

Procedure and Materials

The experiment was performed in two parts. In order to decrease the

effect of social desirability on the behavioral measure of attending a second
appointment, participants were told that the two parts were separate experi-
ments when, in fact, there was a single experiment.
During the first meeting, the experimenter administered the Aitken
Procrastination Inventory (API; Aitken, 1982) and three items measuring
behavioral intention. Following administration of the questionnaires, the
experimenter thanked the students for their participation and explained that
any questions could be directed to her through the psychology department.
Each participant was given extra credit for the class for which they were
participating.
The API (Aitken, 1982) includes 19 items rated on a 5-point Likert scale
ranging from 1 (False) to 5 (True). The ratings are summed after appropriate
reverse-scoring, resulting in a total score ranging from 19 to 95, where higher
scores indicate higher levels of procrastination. The API was developed to
5
Missing data were not a problem for any of the measures in the present study except for
demographic data. There were 3 students who did not report class standing, 4 who did not report
their majors, and 6 who did not report their GPAs.
372 OWENS ET AL.

differentiate chronic procrastinators from nonprocrastinators among college

undergraduates. Cronbach’s alpha (a measure of internal consistency used to
estimate reliability) for both Aitken’s sample and the present sample was .82.
A three-item questionnaire was developed to measure the behavioral
intention variable (see Appendix A). The items were developed according to
the guidelines specified by Ajzen (1985). This methodological issue is dealt
with further in the Discussion. The items were measured using a 7-point scale
ranging from 1 (strongly disagree/unlikely) to 7 (strongly agree/likely). For the
purpose of analysis, the items were summed to create an overall behavioral
intention score. Cronbach’s alpha was found to be .71. API and behavioral
intention were significantly correlated (r = -.47, p < .001).
In the second part of the experiment, the professor in each of the classes
distributed sheets of paper describing an opportunity to earn more extra
credit. The paper listed 10 times during which the participant could report to
a laboratory in the psychology department in order to take part in another
“experiment.” The nonimplementation intention participants received no
further instructions. Implementation intention participants were given the
following instructions along with the list of times:
You are more likely to keep your appointment if you commit
yourself to arriving to the assigned room at one of the times
listed above. Select now the time at which you plan to come for
the second experiment, write it at the bottom of the second
page, and return that page to your instructor.
When participants reported for the second experiment, it was explained to
them that, in fact, there was no second experiment and that this meeting
related back to the questionnaires that they completed in class at the begin-
ning of the semester. Each participant was debriefed thoroughly and given
the opportunity to ask questions. Participants’ presence was noted for extra
credit. Those participants who did not attend the second experiment did not
receive the extra credit.

Results

Procrastination and Implementation Intention

Univariate analyses. There was a statistically significant difference

between implementation intention (II) and no II in behavioral enactment,
thus supporting Hypothesis 1. A majority (47 of 76; 61.8%) of the II partici-
pants enacted the behavior, as opposed to 18.4% (14 of 76) of the no-II
participants, c2(1, N =152) = 29.82, f2 = .20, p < .001.
 OVERCOMING PROCRASTINATION 373

There was also a statistically significant effect for procrastination, thus

supporting Hypothesis 2. Participants were split (Mdn = 50) into low
procrastinators (LP) and high procrastinators (HP).6 LPs evidenced more
behavioral enactment (39 out of 79; 49.4%) than did HPs (22 out of 73;
30.1%). This difference in behavioral rates was statistically significant, c2(1,
N = 152) = 5.84, f2 = .04, p < .016.
Logistic regression analyses. Behavioral enactment was the outcome, and
II and procrastination were the explanatory variables.7 At Step 1, both
explanatory variables were simultaneously entered into the equation.8 At
Step 2, the attribute by treatment interaction (Procrastination ¥ II) was
entered. The interaction was nonsignificant, c2(1, N = 152) = 0.28, p < .598,
and was deleted from the model. The model containing procrastination and
II differed from the intercept-only model, c2(2, N = 152) = 35.30, p < .001,
and is presented in Table 1 (Model I). It can be seen in Table 1 that both
variables’ partial B weights were significant (a < .05) using the Wald and
likelihood ratio tests (see Cohen, Cohen, West, & Aiken, 2003).
The odds ratio of II controlling for procrastination (see Table 1) can be
interpreted as follows: The odds of behavioral enactment were about 7.73
times greater for II participants than for no-II participants.9 The Hosmer–
Lemeshow statistic (Hosmer & Lemeshow, 2000) was used to assess how
adequately Model I fits the data. A good-fitting model will produce a
nonsignificant chi square. As can be seen in Table 1, Model I was found to
provide an acceptable fit to the data ( p < .345).

6
Procrastination was kept as a continuous variable for the logistic regression analyses. It was
dichotomized here to simplify the results and to make them similar in metric (i.e., percentages)
to the chi-square analysis for II effect on behavioral enactment that was just given. Procrasti-
nation was normally distributed; and skewness and kurtosis were nonsignificant (M = 49.59,
SD = 10.84). Scores ranged from 23 to 83.
7
Scaling for the variables was as follows: behavioral enactment, 0 = no enactment, 1 = enact-
ment; implementation intention, 0 = no II, 1 = II. For procrastination, higher scores indicate
more procrastination.
8
This experiment can be conceptualized as an ANCOVA, where II is the treatment manipu-
lation (to which participants are randomly assigned) and procrastination is a continuous pre-
dictor or covariate. The typical ANCOVA analysis assesses a main effect, eliminating the main
effect of the other variable. The analysis here follows the ANCOVA approach used for experi-
mental data. This analysis approach was utilized to test the main effects for all models tested
here.
9
The odds-ratio results are scale-dependent or are not standardized. For this reason, they are
not directly comparable. A 1-unit change for the dichotomous II variable covers the entire range
of that variable; while for the continuous procrastination variable, a 1-unit change covers only
a small section of the variable’s range. Thus, dichotomous variables tend to have larger odds
ratios than do continuous variables. Cohen et al. (2003) recommended some adjustments to
better compare odds ratios across variables. For example, if one could state that a 2-unit change
on procrastination is considered meaningful, then the odds ratio would go from 0.96 to 6.88. In
the present study, it is not clear what a meaningful change in the procrastination metric would
be. For this reason, its odds ratio is not interpreted.
Table 1

Summary of Logistic Regression Models for Understanding Behavioral Enactment

Hosmer–
374 OWENS ET AL.

Measures of fit Lemeshow

Cox–
Likelihood Odds Snell Nagelkerke’s
Variable B SE Walda p ratio (c2)b p ratio RL2 R2 R2 c2 p

Model I PROC -0.04 0.02 3.97* .046 4.21* <.05 0.96 .17 .21 .28 8.97c .345
II 2.05 0.39 27.40* <.001 32.11* <.001 7.73
Intercept 0.30 0.92 0.10 .748 1.34
Model II INT 0.11 0.09 1.48 .224 1.55 <.25 1.11 .18 .22 .29 16.39c* .037
PROC -0.02 0.02 1.19 .276 1.22 <.50 0.98
II 2.00 0.39 25.93* <.001 30.90* <.001 7.40
Intercept -2.31 2.32 0.99 .319
Model III INT 0.16 0.08 4.11* .043 4.54* <.05 1.17 .17 .21 .28 12.71d* .048
II 1.95 0.39 25.55* <.001 29.18* <.001 7.00
Intercept -4.32 1.46 8.74 <.003 0.01

Note. N = 152. PROC = procrastination; II = implementation intention; INT = behavioral intention.

a
df = 1. bdf = 1. cdf = 8. ddf = 6.
*a < .05.
 OVERCOMING PROCRASTINATION 375

An intuitively simpler way of understanding the effects of II controlling

for procrastination is presented in Table 2. In Table 2, the effects of II on
behavioral enactment are examined within low procrastination and high
procrastination. For these analyses—unlike the logistic regression in which
procrastination was kept as a continuous variable—a median split was
applied to the procrastination data to create the low and high groups. As can
be seen in Table 2, II led to a more than 40% increase in behavioral enact-
ment, regardless of whether a participant was high or low on procrastination.

Augmentation and Intention

Univariate analyses. There was a statistically significant relationship

between behavioral enactment and behavioral intention (INT; rpb = .20,
p < .014). The more a person intended to perform the behavior, the more
likely it was that the behavior was performed, thus supporting Hypothesis 3.
The INT scale ranged from 3 to 21, where 3 represents no intention and 21
represents a complete intention to perform the behavior. INT was leptokurtic
and negatively skewed (statistically significant at a < .05; M = 18.18,
SD = 2.92). In effect, the majority of participants fully intended to engage in
the behavior. Indeed, 68.4% (104 out of 152) scored 18 or higher on INT.
There were 47 participants (out of 152; 30.9%) who had the highest score of
21 on the INT. Yet, only 40.1% (61 out of 152) actually engaged in the
behavior. Clearly, while INT does predict behavioral enactment, it does not
give a complete understanding of what brings about behavior.
Logistic regression analyses. To assess whether II (and procrastination)
augment the effects of INT, the following model was tested: At Step 1, all
three explanatory variables were entered simultaneously into the equation for
understanding behavioral enactment. At Step 2, the Attribute ¥ Attribute
and Attribute ¥ Treatment interactions were entered. The interactions were
not statistically significant, c2(4, N = 152) = 6.32, p < .177, and were deleted
from the equation. The model containing INT, procrastination, and II dif-
fered significantly from the intercept-only model, c2(3, N = 152) = 36.85,
p < .001, and is presented in Table 1 (Model II). Using the Wald and likeli-
hood ratio tests, it can be seen in Table 1 that only the partial B weight for II
was statistically significant.
One explanation for the nonsignificant partial B weights for INT and
procrastination is that of collinearity. The correlation between them was
statistically significant (r = -.51, p < .001). In order to determine if these two
measures are conceptually distinct, a factor analysis was performed. Specifi-
cally, the API items and the INT items were subjected to a principal axis
factor analysis with an oblique (direct oblimin) rotation (see Appendix B).
 376 OWENS ET AL.

Table 2

Frequencies of Behavioral Enactment as a Function of Implementation Intention Within High and Low Procrastination

High procrastination Low procrastination

II No II Difference II No II Difference
Behavior 19 51.4% 3 8.3% 41.3% Behavior 28 71.8% 11 27.5% 44.3%
No behavior 18 48.6% 33 91.7% No behavior 11 28.2% 29 72.5%
Total 37 36 Total 39 40
2 2
c (1, N = 73) = 16.04*, f2 = .22, p < .001 c (1, N = 79) = 15.50*, f2 = .20, p < .001
Note. II = implementation intention.
*a < .05.
 OVERCOMING PROCRASTINATION 377

Those results are supportive of the notion that API and INT items are
measuring different constructs. That is, four factors underlie the API and
another factor underlies the behavioral intention items. Indeed, only one item
from the API loaded onto the INT factor, and no behavioral intention items
loaded onto any of the API factors.
The adequacy of fit of Model II to the data was not acceptable (see Table 1;
Hosmer–Lemeshow result, p < .037; Hypothesis 4). This result led to one last
set of analyses. Specifically, procrastination was dropped from the model, and
the effects of INT and II on behavioral enactment were assessed.
At Step 1, both INT and II were simultaneously entered into the equation.
At Step 2, the Attribute ¥ Treatment interaction (INT ¥ II) was entered.
The interaction was nonsignificant, c2(1, N = 152) = 0.73, p < .391, and was
dropped from the model. The model containing INT and II differed from the
intercept-only model, c2(2, N = 152) = 35.63, p < .001, and is presented in
Table 1 (Model III). While both partial B weights were statistically significant
(see Table 1), the model does not adequately fit the data, as demonstrated by
the statistically significant Hosmer–Lemeshow test (see Table 1).
Research by Sheeran et al. (2005) supports the idea of an interaction
between goal intentions and II such that strong or activated goal intentions
and II can be most effective in bringing about behavioral initiation (Hypoth-
esis 5). Given that an interaction did not emerge in the present study, we
further investigated these two variables. Participants were divided into two
groups based on their intention regarding keeping appointments. A median
split was conducted, and participants were classified as lower intention
(n = 74) or higher intention (n = 78). The efficacy of II was assessed for these
two groups separately. These results are presented in Table 3.
As can be seen in Table 3, lower intention/no-II participants kept their
appointments 10.8% of the time, while lower intention/II participants kept
their appointments 48.6% of the time. The difference between these two
groups is statistically significant ( p < .001). For higher intention participants
(see Table 3), the pattern was the same. The no-II participants kept the
appointments less often than did the II participants (25.6% vs. 74.4%). While
II was always beneficial, the effect was somewhat stronger for higher inten-
tion participants (f2 = .24) than for the lower intention participants (f2 = .17).
This finding is consistent with Sheeran et al. (2005).
The reason for the nonsignificant interaction is plausibly explained in at
least two ways. First, as mentioned earlier, the majority of participants had a
very high level of intention; thus, few respondents had no intention at all.
Second, the level of specificity for intention, as a result of methodological
issues, was not at the recommended level (i.e., keeping appointments vs.
keeping this particular appointment; see Appendix A). These issues are dealt
with further in the Discussion.
Table 3
378 OWENS ET AL.

Frequencies of Behavioral Enactment as a Function of Implementation Intention Within Lower and Higher Intention
Groups

Lower intention Higher intention

II No II Difference II No II Difference
Behavior 18 48.6% 4 10.8% 37.8% Behavior 29 74.4% 10 25.6% 48.8%
No behavior 19 51.4% 33 89.2% No behavior 10 25.6% 29 74.4%
Total 37 37 Total 39 39
c2(1, N = 74) = 12.68*, f2 = .17, p < .001 c2(1, N = 78) = 18.51*, f2 = .24, p < .001
Note. II = implementation intention.
*a < .05.
 OVERCOMING PROCRASTINATION 379

In examining the three models presented in Table 1, it can be seen that

only Model I had an acceptable fit to the data. Furthermore, the R2 analogs
for logistic regression presented in Table 1 (RL2, Cox–Snell R2, and
Nagelkerke’s R2) all indicate that all three models yield a similar gain in
prediction of behavioral enactment. The more complicated Model II gives no
real increment in prediction over the more parsimonious Models I and III.
Given that only Model I exhibited an acceptable fit to the data, as assessed by
the Hosmer–Lemeshow test, that model is endorsed over the others.

Discussion

Procrastination had a clear effect on whether or not participants kept

their appointments. Those who rated themselves as low procrastinators (LPs)
kept their appointments more often than did those who rated themselves as
high procrastinators (HPs). In addition, those who formed implementation
intentions (IIs) were nearly 8 times more likely to keep their appointments
than were those who did not form IIs. While the effects for IIs within HPs and
LPs were nearly the same, it is evident that there was an improvement in
behavioral enactment for those HPs who formed IIs. Thus, both HPs and
LPs benefited from forming implementation intentions.
There was a significant association between behavioral intentions and
behavioral enactment. Those with stronger intentions to behave were more
likely to keep one of the offered extra-credit appointments.
With regard to a model of procrastination, the results indicate that the
best prediction of behavioral enactment includes main effects for procrasti-
nation and implementation intentions. As suggested by some (e.g., Van
Eerde, 2000), procrastination occurs when strong intentions are not trans-
lated into action. Given the significant association between behavioral inten-
tions and behavioral enactment and research by Sheeran et al. (2005), it is
interesting that there was no significant interaction between behavioral inten-
tions and implementation intentions. It was hypothesized that this would be
a result either of the fact that most participants have strong intentions, or a
methodological issue that precluded the experimenters from measuring inten-
tion to attend this particular appointment.10
While it is important to measure intentions related to a specific behavior,
the measure of behavioral intentions in the current experiment was worded so
that participants would not be able to make the connection between two
supposedly separate studies. In future experiments examining behavioral
intentions and procrastination, this issue will be of paramount importance.
Despite the lack of a significant interaction between behavioral intentions and

10
We thank a reviewer for bringing this issue to our attention.
380 OWENS ET AL.

implementation intentions, the current results indicate that the effect of

implementation intentions was somewhat stronger in participants with higher
behavioral intentions than in participants with lower behavioral intentions.

References

Aitken, M. E. (1982). A personality profile of the college student procrastina-

tor. Unpublished doctoral dissertation, University of Pittsburgh.
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In
J. Kuhl & J. Beckman (Eds.), Action control: From cognition to behavior
(pp. 11–39). Berlin, Germany: Springer-Verlag.
Ajzen, I. (1991). The theory of planned behavior. Organizational and Human
Decision Processes, 50, 179–211.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behavior:
Attitudes, intentions, and perceived behavioral control. Journal of Experi-
mental Social Psychology, 22, 453–474.
Brandstätter, V., Lengfelder, A., & Gollwitzer, P. M. (2001). Implementation
intentions and efficient action initiation. Journal of Personality and Social
Psychology, 81, 946–960.
Burka, J. B., & Yuen, L. M. (1983). Procrastination: Why you do it and what
to do about it. Reading, MA: Addison-Wesley.
Chu, A. H. C., & Choi, J. N. (2005). Rethinking procrastination: Positive
effects of “active” procrastination behavior on attitudes and perfor-
mance. Journal of Social Psychology, 145, 245–264.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple
regression/correlation analysis for the behavioral sciences (3rd ed.). Hills-
dale, NJ: Lawrence Erlbaum.
Day, V., Mensink, D., & O’Sullivan, M. (2000). Patterns of academic pro-
crastination. Journal of College Reading and Learning, 30, 120–134.
Dill, C. A., Owens, S. G., Homburger, L. A., Travers, M., & Lancaster, J. M.
(2002, August). Initiating/maintaining behavior: Impact of reinforcement,
goal, and implementation intentions. Presented at the annual meeting of
the American Psychological Association, Chicago, IL.
Ellis, A., & Knaus, W. (1977). Overcoming procrastination. New York: Insti-
tute for Rational Living.
Ferrari, J. R. (2001). Getting things done on time: Conquering procrastina-
tion. In C. R. Snyder (Ed.), Coping with stress: Effective people and
processes (pp. 30–46). Oxford, UK: Oxford University Press.
Ferrari, J. R., & Tice, D. M. (2000). Procrastination as a self-handicap for
men and women: A task-avoidance strategy in a laboratory setting.
Journal of Research in Personality, 34, 73–83.
 OVERCOMING PROCRASTINATION 381

Flett, G. L., Blankstein, K. R., & Martin, T. R. (1995). Procrastination,

negative self-evaluation, and stress in depression and anxiety: A review
and preliminary model. In J. R. Ferrari & J. L. Johnson (Eds.), Procras-
tination and task avoidance: Theory, research, and treatment (pp. 137–
167). New York: Plenum.
Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple
plans. American Psychologist, 54, 493–503.
Gollwitzer, P. M., & Brandstätter, V. (1997). Implementation intentions and
effective goal pursuit. Journal of Personality and Social Psychology, 73,
186–199.
Gollwitzer, P. M., & Sheeran, P. (in press). Implementation intentions and
goal achievement: A meta-analysis of effects and processes. Advances in
Experimental Social Psychology.
Harriott, J., & Ferrari, J. R. (1996). Prevalence of procrastination among
samples of adults. Psychological Reports, 78, 611–616.
Holland, T. (2001). The perils of procrastination. Far Eastern Economic
Review, 164, 66–72.
Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression (2nd ed.).
New York: Wiley.
Kegley, C. W. (1989). The Bush administration and the future of American
foreign policy: Pragmatism or procrastination? Presidential Studies Quar-
terly, 19, 717–731.
Knaus, W. (2000). Procrastination, blame, and change. Journal of Social
Behavior and Personality, 15, 153–166.
Koestner, R., Lekes, N., Powers, T. A., & Chicoine, E. (2002). Attaining
personal goals: Self-concordance plus implementation intentions equals
success. Journal of Personality and Social Psychology, 83, 231–244.
Koole, S., & van’t Spijker, M. (2000). Overcoming the planning fallacy
through willpower: Effects of implementation intentions on actual and
predicted task-completion times. European Journal of Social Psychology,
30, 873–888.
Milgram, N. A., Mey-Tal, G., & Levison, Y. (1998). Procrastination, gener-
alized or specific, in college students and their parents. Personality and
Individual Differences, 25, 297–316.
Orbell, S., Hodgkins, S., & Sheeran, P. (1997). Implementation intentions
and the theory of planned behavior. Personality and Social Psychology
Bulletin, 23, 945–954.
Orbell, S., & Sheeran, P. (2000). Motivational and volitional processes in
action: A field study of the role of implementation intentions. Journal of
Applied Social Psychology, 30, 780–797.
Owens, S. G. (2003). The moderating effect of stress on implementation
intentions and the effect of reinforcement on number of arguments
382 OWENS ET AL.

regarding capital punishment. Unpublished doctoral dissertation, Hofstra

University.
Owens, S. G., Dill, C., Levine, N., & Goldfarb, L. (2001, January). Do goal
benefits and task interference mediate the effect of implementation inten-
tions on goal completion? Presented at the 2nd annual meeting of the
Society for Personality and Social Psychology, San Antonio, TX.
Owens, S. G., Dill, C., Levine, N., & Goldfarb, L. (2002, August). Examining
the effectiveness of implementation intentions: Task features and auto-
maticity. Presented at the annual meeting of the American Psychological
Association, Chicago, IL.
Potts, T. (1987). Predicting procrastination on academic tasks with self-report
measures. Unpublished doctoral dissertation, Hofstra University.
Pychyl, T. A., Lee, J. M., Thibodeau, R., & Blunt, A. (2000). Five days
of emotion: An experience sampling study of undergraduate student
procrastination. Journal of Social Behavior and Personality, 15(5),
3–13.
Schouwenburg, H. C. (2004). Procrastination in academic settings:
General introduction. In H. C. Schouwenburg, C. H. Lay, T. A.
Pychyl, & J. R. Ferrari (Eds.), Counseling the procrastinator in
academic settings (pp. 3–18). Washington, DC: American Psychological
Association.
Schouwenburg, H. C., Lay, C. H., Pychyl, T. A., & Ferrari, J. R. (Eds.).
(2004). Counseling the procrastinator in academic settings. Washington,
DC: American Psychological Association.
Sheeran, P. (2002). Intention–behavior relations: A conceptual and empirical
review. In W. Stroebe & M. Hewstone (Eds.), European review of social
psychology (Vol. 12, pp. 1–30). New York: Wiley.
Sheeran, P., & Orbell, S. (1999). Implementation intentions and repeated
behavior: Augmenting the predictive validity of the theory of planned
behavior. European Journal of Social Psychology, 29, 349–369.
Sheeran, P., Webb, T. L., & Gollwitzer, P. M. (2005). The interplay between
goal intentions and implementation intentions. Personality and Social
Psychology Bulletin, 31, 87–98.
Sirois, F. M. (2004). Procrastination and intentions to perform health be-
haviors: The role of self-efficacy and the consideration of future
consequences. Personality and Individual Differences, 37, 115–128.
Sirois, F. M., Melia-Gordon, M. L., & Pychyl, T. A. (2003). “I’ll look after
my health later”: An investigation of procrastination and health. Person-
ality and Individual Differences, 35, 1167–1184.
Solomon, L. J., & Rothblum, E. D. (1984). Academic procrastination:
Frequency and cognitive-behavioral correlates. Journal of Counseling
Psychology, 31, 503–509.
 OVERCOMING PROCRASTINATION 383

Steel, P., Brothen, T., & Wambach, C. (2001). Procrastination and person-
ality, performance, and mood. Personality and Individual Differences, 30,
95–106.
Tice, D. M., & Baumeister, R. F. (1997). Longitudinal study of procrastina-
tion, performance, stress, and health: The costs and benefits of dawdling.
Psychological Science, 8, 454–458.
Van Eerde, W. (2000). Procrastination: Self-regulation in initiating aversive
goals. Applied Psychology, 49, 372–389.
Van Eerde, W. (2003). A meta-analytically derived nomological network of
procrastination. Personality and Individual Differences, 35, 1401–1418.
Verplanken, B., & Faes, S. (1999). Good intentions, bad habits, and effects of
forming implementation intentions on healthy eating. European Journal
of Social Psychology, 29, 591–604.

Appendix A

Behavioral Intention Variable

1. I will keep appointments.
Strongly agree 1 2 3 4 5 6 7 Strongly disagree
2. I intend to keep appointments.
Strongly agree 1 2 3 4 5 6 7 Strongly disagree
3. How likely or unlikely is it that you will keep appointments?
Unlikely 1 2 3 4 5 6 7 Likely

Appendix B

Pattern Loading and Factor Correlation Matrices for Factor Analytic

Results of API and INT Measures
Inventory Factor

Delay of Missed Missed Recognition Behavioral

API INT task deadline appointment of problem intention

Item 1 .60
Item 3 .62
Item 7 .32
Item 9 .46
384 OWENS ET AL.

Appendix B

Continued
Inventory Factor

Delay of Missed Missed Recognition Behavioral

API INT task deadline appointment of problem intention

Item 10 .81
Item 11 .67
Item 12 .74
Item 14 .39 .39
Item 15 .73
Item 2 .50
Item 4 .50
Item 19 .34
Item 6 .78
Item 16 .43
Item 17 .68
Item 18 .66
Item 5 -.37
Item 8 -.50 -.35
Item 1 .81
Item 2 .50
Item 3 .30

Variable 1 2 3 4

1. Delay —
2. Missed deadline .37 —
3. Missed appointment .23 .13 —
4. Recognition of problem -.01 -.15 -.01 —
5. Behavioral intention -.42 -.24 -.28 .16

Note. N = 152. API = Aitken Procrastination Inventory (Aitken, 1982). INT = behavioral intention.
The number of factors retained was based on the scree test and eigenvalues greater than 1 rule. The
principal axis factor results were subjected to a direct oblimin (oblique) rotation (d = -.50). For
convenience, items have been grouped by the factor on which they loaded. While all factor correlations
are reported, only pattern loadings of |.30| or greater are reported. Only API Item 13 failed to load on
any of the factors.