Factors Affecting Procrastination

There is evidence that procrastination is caused and maintained by a variety of factors, which include the lack of intention building (Owens et al., 2008), poor planning and time management (Lay and Schouwenburg, 1993; Specter and Ferrari, 2000), and time discounting (Howell et al., 2006). Time discounting refers to the tendency individuals have to discount future reward depending on the interval between the activity and reward. A greater interval reduces the motivational value of the reward (Steel and König, 2006). In addition, affective obstacles (Sirois, 2014) like distress (Pychyl et al., 2000; Tice et al., 2001), anxiety (Rothblum et al., 1986), and low positive affect (Ferrari and Díaz-Morales, 2007; Sirois and Pychyl, 2013) mislead individuals into procrastinating to “repair” their mood. Moreover, Wohl et al. (2010) found that forgiving oneself for procrastinating over a specific task decreases subsequent procrastination by decreasing negative affect. Negative affect increases rumination on the transgression of procrastination, and vice versa (Thompson et al., 2005). To achieve short-term mood repair, individuals seek pleasant distractions and postpone tasks they should be doing. But forgiveness “allows the individual to move past their maladaptive behavior” (Wohl et al., 2010, p. 806) and focus on the next task. Their dysfunctional need for procrastination to repair their mood seems to be reduced. Consistent with this, Eckert et al. (2016) showed that one’s ability to cope adaptively with aversive emotions reduces one’s subsequent likelihood of procrastination.

Another equally-important factor that is believed to increase procrastination is a lack of self-reinforcement (e.g., Ferrari and Emmons, 1995). If tasks are boring or fail to induce a positive affect, individuals with procrastination problems tend to seek pleasant distractions. However, individuals who reinforce themselves for doing the boring or aversive task tend to report less procrastination (Ferrari and Emmons, 1995). Consequently, improving these aforementioned factors should subsequently decrease procrastination.

It can therefore be summarized that (1) lack of intention building, (2) lack of planning and time management, (3) difficulties initiating and maintaining a certain action caused by emotional obstacles, poor mood or delayed gratification, and (4) the absence of self-reinforcement and negative self-evaluations, including low self-efficacy expectations, all are associated with procrastination.

Interventions to Reduce Procrastination

Due to the fact that procrastination causes a lot of problems (e.g., academic impairment, financial problems), interventions to overcome or reduce procrastination are needed. Regarding behavioral interventions, stimulus control provides strategies removing aspects that might interfere with the task (Mulry et al., 1994). In his meta-analysis, Steel (2007) found that interventions that foster automaticity reduced procrastination. Because procrastination is related to avoiding behavior, gradually exposing individuals to aversive activities seems to reduce procrastination (Brown, 1991). In order to overcome lack of commitment with the task, adequate goal setting increases motivation (Boice, 1989).

Since irrational believes and attitudes, like perfectionism, fear of failure, and self-doubt, promote procrastination cognitive interventions, like restructuring, raising self-esteem, and behavioral experiments that facilitate corrective experiences, may reduce procrastination (Rozental and Carlbring, 2014). Glick and Orsillo (2015) found effects of acceptance-based behavioral therapy on procrastination.

Few clinical trials have examined the efficacy of treatment interventions for procrastination (Rozental and Carlbring, 2014). For example, Höcker et al. (2013) reduced procrastination using intervention modules that focused on starting tasks on time, formulating realistic plans of action, and restricting working times. In student samples interventions containing planning and time-management (Schmitz and Wiese, 2006; Häfner et al., 2014) as well as self-instruction methods, such as to stop negative thoughts or positive self-talk (Schmitz and Wiese, 2006), reduced procrastination significantly. In contrast to face-to-face interventions, internet-based interventions are cost-effective in the treatment of a wide range of problems (Rozental et al., 2015). Surprisingly, only a few randomized controlled trial (RCT) have investigated internet-based interventions in this field (Wäschle et al., 2014; Rozental et al., 2015), clearly justifying the need for further RCTs of Internet-based interventions to decrease procrastination.

Rozental et al. (2014) found that a 10-week internet-based intervention providing psychoeducation and several techniques in 10 modules reduced procrastination with a medium effect size. The techniques included behavioral activation, graded exposure, behavioral experiments, identifying and testing rigid beliefs and assumptions, and stimulus control. In a three-armed randomized controlled study, 150 participants were randomized either on treatments with therapist contact (guided), treatment without therapist contact (self-guided), or wait-list control. Compared with the wait-list control, the guided condition revealed greater effect sizes than the self-guided condition.

First Purpose of the Present Study

Internet-based interventions provide many advantages like improved access, cost-effectiveness, local and time independence. Thus, the first purpose of the present study is to develop and to evaluate a brief self-guided internet-based intervention to reduce procrastination utilizing the advantages. In the following section we describe factors affecting procrastination. Based on these factors, we provide the Rubicon-model as heuristic for the development of an intervention. But internet-based interventions have also disadvantages like decreased adherence compared to face-to-face treatments (Richards and Richardson, 2012; Andersson and Titov, 2014). That followed, we describe factors affection adherence and we suggest text messages (SMS) as intervention to reduce the problems with adherence in internet-based interventions. Thus, the second purpose of the present study is to investigate whether additional SMS support increased adherence to (as well as efficacy of) the internet-based intervention. In contrast to the intervention of Rozental et al. (2014), the present study investigated the effectiveness of a 2-week intervention which provides SMS support instead of therapist contact. With regard to the content, the Online-Training to Overcome Procrastination (ON.TOP) is orientated on the Rubicon-model from Heckhausen and Gollwitzer (1987) as shown below.

The Rubicon-Model as a Heuristic for Developing an Intervention

Addressing factors that cause and maintain procrastination, the Rubicon-model (Heckhausen and Gollwitzer, 1987) was used as a heuristic for developing interventions to reduce procrastination. The Rubicon-model is a theory of action regulation (Heckhausen and Gollwitzer, 1987) that differentiates four motivational and volitional phases: (1) intention-building; (2) time-management planning and realistic goal setting; (3) shielding the intended action from distractions; and (4) evaluating the process and the results of completed activities. The first phase is pre-decisional. It emphases the process of pondering the “pros and cons of one’s wishes […] by assessing the desirability of expected outcomes and the question of feasibility” (Achtziger and Gollwitzer, 2007, p. 769). This leads into constructing intentions to act. For the purpose of acting successful, the pre-actional second phase focuses on planning when, where and how one must work toward the goal. It includes several volitional processes. In the actional third phase, goal-directed behaviors must be initiated and maintained by volitional processes like emotion regulation. Upon completion of goal-directed behaviors, during the post-actional fourth phase, the outcome must be evaluated. This evaluation influences future intention building by influencing self-efficacy and action-outcome expectations. This heuristic model addresses factors that both cause and maintain procrastination; as such, it provided the basis for developing our intervention to reduce procrastination.

Factors Affecting Adherence to Online-Based Interventions

Adherence to treatment is defined as the extent to which the participant of an intervention coincides with the prescribed treatment (Urquhart, 1996). Research has identified certain factors that increase adherence to internet-based interventions. These include tailored feedback by e-coaches (Cugelman et al., 2011); program interactivity (Hurling et al., 2006); an enriched training environment that includes features like multimedia presentations and audio-exercises (Webb et al., 2010); and text messages as reminders (Fjeldsoe et al., 2009; Krishna et al., 2009). In particular, short-message service (SMS) support appears to enhance the effectiveness of internet-based health interventions (Webb et al., 2010), having been applied in the research literature to fulfill a variety of functions. For example, Kamal et al. (2015) increased medical adherence among stroke patients by providing reminders and health information by SMS. Similarly, Koshy et al. (2008) reminded ophthalmology outpatients of appointments and, thereby, increased their adherence. Kaptein et al. (2012) implemented successfully social influence strategies as prompts, via SMS, in order to reduce snacking. However, the success of this strategy seems to depend upon individually-tailored messages, with the effectiveness of non-tailored messages less apparent.

Another reason SMS increases adherence may be seen with the so-called “foot-in-the-door” (FITD) technique. This technique works by first inducing a “yes” response for a small request. This primary “yes” increases the probability of receiving an affirmative response to subsequent, greater requests. Several investigators have found evidence that the FITD technique influences behavior (e.g., Freedman and Fraser, 1966; Guéguen and Fischer-Lokou, 1999; Guéguen et al., 2008), and it appears to be an effective strategy for real-world interventions (Guéguen, 2002; Cugelman et al., 2009; Grassini et al., 2013). As such, to increase adherence, very small exercises provided daily via SMS may be appropriate. These mini-exercises may be the proverbial “foot in the door” and increase the probability of later engagement with the intervention. However, to the best of our knowledge, no prior research has systematically investigated whether the FITD technique has any role increasing adherence through SMS support.

Goals of the Present Study

The present study was designed to evaluate the internet-based intervention ON.TOP, which was developed in accordance with the Rubicon model (Heckhausen and Gollwitzer, 1987). Relative to the first Internet-based intervention created to reduce procrastination, reported by Rozental et al. (2015), ON.TOP focuses more on strategies to cope adaptively with emotional obstacles cueing procrastination; for example, strategies to tolerate and modify aversive emotions (Eckert et al., 2016). In this way, the current study will enrich the research literature on Internet-based interventions targeting procrastination.

Recalling that Internet-based interventions often must confront problematic adherence, a second aim of this study was to investigate if daily SMS support increases adherence to and/or the efficacy of the intervention. Thus, every day, participants in one of the two intervention arms of the study received two short exercises via text messages (SMS support). These exercises were intended to motivate subjects in various ways, including use of the foot-in-the-door technique by which affirmative responses to simple requests were used as a springboard to further engagement in anti-procrastination training exercises.

Procedures and the Sample

Since half of students report serious procrastination problems (Day et al., 2000), we decided to recruit subjects from a student population. To recruit such students, we distributed information (a) via the internal communication system at Leuphana University of Lüneburg, (b) via several helplines for students at three German universities (in Hannover, Hildesheim, and Lueneburg), and (c) via the Moodle Communication System used by the co university in Hagen (Germany). All participants reported to be university students. All individuals completing the baseline online survey (T0) and providing informed consent were included. After completing the baseline survey, participants were randomized in Excel, using the RandBetween function, which automatically assigned the number 0, 1, or 2 to each ID number, indicating allocation to the I_A, I_SMS, or WLC group, respectively.

Participants who were randomized to receive either I_A or I_SMS received access to the internet-training program via e-mail. Additionally, subjects in the I_SMS group also received two text messages daily (SMS support). Waiting list controls merely received information about the progress of the study. Two weeks after completing the baseline questionnaire, and immediately after completion of the 2-week intervention, all participants were asked to complete a second questionnaire (T1), followed by a final follow-up questionnaire 8 weeks after completing the second questionnaire. At that point, subjects in the WLC group were granted access to the ON.TOP program. At all three data collection points (T0, T1, T2) participants completed the General Procrastination Scale (Klingsieck and Fries, 2012). At baseline, socio-demographic data also were collected; at the immediate post-treatment assessment (T1), subjects were asked to indicate the frequency of their engagement with the training program over the preceding 2 weeks. Figure ​Figure11 shows the flow diagram of this study.

Open in a separate window

FIGURE 1

CONSORT flow diagram.

A total of 161 students were randomized to either the I_A (N = 58; 67.2% females), the I_SMS (N = 55; 74.5% females), or WLC group (N = 48; 81.3% females). Overall, 119 of the subjects (73.9%) were women, and the average age was 28.4 years (SD = 8.9), ranging from 19 to 62 years; this broad age range is explained by our inclusion of several mature students who had returned to school later in life. Regarding the sample’s career choice, 63 participants (39.1%) studied educational science or related subjects, 42 (26.1%) studied psychology, 12 (7.5%) studied economic or related science, 7 (4.3%) studied engineering, 5 (3.1%) studied cultural studies, 3 (1.9%) studied politics, 3 (1.9%) studied informatics, 3 (1.9%) studied environmental sciences, 3 (1.9%) studied digital media and another studied jurisprudence (0.6%). Nineteen participants (11.8%) did not report their subject. Twenty-seven (16.8%), five (3.1%), and three (1.2%) of the 161 subjects stated that they had been diagnosed with depression, anxiety, or attention deficit-hyperactivity disorder (ADHD), respectively. No differences in the three treatment groups were apparent for gender distribution (χ² = 2.69; p = 0.26) or age (F = 1.049; p = 0.35). Similarly, there were no inter-group differences in the percentage with depression (χ² = 0.979; p = 0.61), anxiety (χ² = 3.488; p = 0.18), or ADHD (χ² = 0.019; p = 0.99).

Intervention

ON.TOP combines already-available, well-established therapeutic techniques to reduce procrastination. It consists of four sessions which includes videos, audio exercises (relaxation or imagination exercises), and written material. To reduce potential obstacles to starting or maintaining the intervention, (1) all text-based information was also provided in audio format, and (2) no session exceeded 30 min in duration. To minimize time-discounting effects, participants were asked to reward themselves with some form of positive reinforcement every evening for every successful attempt they had made that day to decrease procrastination, no matter how small the attempt. Figure ​Figure22 display the timetable of the sessions.

Open in a separate window

FIGURE 2

Timetable of the ON.TOP sessions.

The Rubicon Model (Heckhausen and Gollwitzer, 1987) was applied as the theoretical framework. It includes the following phases: (a) motivational phase (intention building), (b) pre-actional phase (planning), (c) actional phase (realizing), and (d) post-actional phase (evaluating). The 2-week intervention was therefore comprised of four sessions.

During the first session (intention building), participants learned about the relevance of decision-making (Steel, 2007) by watching a psychoeducation video. Keller et al. (2011) found that enhancing active choice fosters desired behaviors. Thus, in the first session, participants were trained to actively decide whether to implement an intention or not. To achieve this, they were asked to identify one of their daily tasks that they were most likely to delay completing. To generate the intention to complete this task, they then were asked to contrast the long-term benefits of executing the task against the long-term costs of avoiding or postponing the task (e.g., failing an examination). They also were asked to compare the short-term costs of executing the task (e.g., boredom) and short-term benefits of avoiding or postponing it. Finally, once they had determined if there were any good reasons why the task should be postponed or avoided, participants were invited to actively decide to execute the task and accept the short-term costs and protocol them in a diary. The diary was provided in PDF format. Decision- and goal setting strategies are often considered suitable for addressing problems of procrastination from cognitive behavior therapy (CBT) (Steel, 2007).

In the second session (planning), participants learned about two principles of planning. The first of these is realistic goal setting, achieved by establishing concrete intermediate goals, a process that tends to increase one’s probability of executing the ultimately-intended task (van Eerde, 2000). One example of this is suggesting that students plan to read and summarize a single chapter per day, instead of studying until they can no longer continue. The second principle involves implementing simple “if-then” plans to determine when and how to execute the task and, thereby, reduce the gap between intentions and behaviors (Gollwitzer, 1999). Previous research indicates that applying implementation intentions may reduce procrastination (van Hooft et al., 2005; Gollwitzer and Sheeran, 2006). Like the first session, subjects were asked to identify one of their daily tasks which they were most likely to procrastinate against doing. If they decided to actively execute the task, they were encouraged to plan task execution by setting up realistic (sub-) goals and applying if-then plans and protocoling them in a PDF diary.

In the third session (realization), subjects learned how to overcome affective obstacles that created gaps between their intentions and behaviors (Eckert et al., 2015). Eckert et al. (2016) have shown that two emotion-regulation strategies are effective at reducing procrastination: tolerating aversive emotions and modifying them. The third session of the ON.TOP program focused on these two strategies. As in the previous sessions, participants were asked to choose one daily task they were most likely to try to delay completing, to actively decide whether to execute the task, and then to plan task execution. They then were asked to (a) identify and label which aversive emotions were cued by the task, then (b) tolerate, and finally (c) modify the aversive emotion. As per Berking and Whitley (2014), the strategy to tolerate aversive emotions included intentionally permitting aversive emotions to be present, reminding oneself of one’s toughness and resilience, and finally reminding oneself of (or increasing) one’s affective commitment with the task. The modification strategy of aversive emotions consisted of first practicing a short relaxation exercise, then reappraising the harm and probability of the potential threat, and lastly deciding whether to execute the task. After completing the chosen task, participants evaluated how successfully they had coped with their aversive emotions to increase their emotional self-efficacy. The emotion regulation strategies were presented via audio files.

During the fourth and final session (evaluating), expectations of self-efficacy were fostered, as self-efficacy expectations are a relevant negative predictor of procrastination (Steel, 2007). Hence, participants evaluated situations in which they were able to reduce procrastination successfully (Sirois, 2004). This may increase both self-efficacy expectations and self-reinforcement to subsequently reduce procrastination. Additionally, they were asked to reflect on why previous strategies had been unsuccessful, so they might work to improve or replace them.

SMS Support

Subjects in the I_SMS group received two brief messages each day, all delivered via SMS on their own cellphone. Each message gave them a short, simple exercise or task to complete, none requiring more than 30 s. The objective of these exercises always was aligned with the content of the current intervention session. Examples of SMS content are: “Which task are you most likely going to postpone today? What are the consequences if you decide not to do the task? Are there any good reasons against this decision?” (Session 1). “Which task are you most likely going to postpone today? If you have decided to do this task, set the time when you will begin now” (Session 2). “Remember a situation in which you overcame procrastination successfully. Try to remember this feeling of success before you will start another aversive task” (Session 3). “In this training program, you have learned how to overcome procrastination. Now it is time to look back and acknowledge your own success” (Session 4). Therefore, participants in both intervention groups followed a timetable that indicated to them when to start and when to end each session (Session 1: day 1–3; Session 2: day 4–7; Session 3: day 8–12, and Session 4: day 13 and 14). Participants in the I_SMS arm received reminders to start their next session via SMS.

In each of the examples listed above, note how every SMS message contains a small, simple task or exercise (e.g., set a time), which motivates participants to complete further training. The threshold for complying is very low. Using this foot-in-the-door (FITD) technique, each “yes” to a small request was hypothesized to increase the probability that the subject would be adherent and engage in later, more time-consuming exercises.

Measurements

Treatment Efficacy

These results based on intention-to-treat analyses (ITT-analyses). In order to compare the three treatment arms simultaneously, the means of the GPS of each point of measurement as well as the group factor (WLC; I_A versus I_SMS) were included in an ANOVA with repeated measures. The main effect of measurement time (F_4,316 = 54.71, p < 0.001, ${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$ = 0.415) as well as the interaction term ‘time × condition’ was significant (F_4,316 = 7.91; p < 0.001, ${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$ = 0.110) were significant. Means for the GPS decreased over time in all three groups. Means and standard deviations for both ITT-Analysis and Completer-analyses are displayed in Table ​Table11. Post hoc contrast at T1 revealed no significant differences between the groups. At T2, Dunnett-T indicated significant differences between WLC and I_SMS, but not between WLC and I_A (see Table ​Table22).

Effect sizes for I_SMS seemed to be stronger than effect sizes for I_A. Comparing WLC against the two intervention groups at T2, we calculated a small effect size for I_A, for GPS (d = 0.29; 95% CI [-0.703, 0.063]), but a medium effect size for I_SMS, again for the GPS (d = 0.57; 95% CI [-0.964, -0.182]). Calculating effect sizes within groups, procrastination seemed to be more reduced by I_SMS (d = 1.15; 95% CI [-1.719, -0.578]) than by I_A (d = 0.85; 95% CI [-1.390, -0.315]). Figure ​Figure33 reveals changes in procrastination as measured on the GPS between T0 and T2.

Open in a separate window

FIGURE 3

Changes in the extent of procrastination (GPS) from pre-measurement (T0) to post-measurement (T1) to 8-week follow-up (T2) in the intervention groups (I_A and I_SMS) and the waiting list control (WLC) group.

In order to investigate if additional SMS-support makes a difference, we comparing both intervention groups directly against each other (I_A versus I_SMS). A 2 × 3 ANOVA (with repeated measures) revealed a significant interaction term between group and time (F_2,111 = 3.67, p < 0.05, ${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$ = 0.062).

Regarding effectiveness, when only completers were included, group differences were still significant (see Table ​Table11). Effect sizes at the time of final follow-up (T2) were similar to those we identified on intention-to-treat-analyses: I_SMS: d = 0.87; 95% CI [-1.361, -0.393]; I_A: d = 0.24; [-0.706, 0.222]; d = 0.28; [-0.745, 0.184]. Moreover, comparing the efficacy of both intervention groups directly via additional ANOVA with repeated measures, the interaction term ‘time × condition’ was significant for GPS (F_2,63 = 5.31; p < 0.01, ${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$ = 0.144).