Moving motivation Stolte, E.
2018
document version
Publisher's PDF, also known as Version of record
Link to publication in VU Research Portal
citation for published version (APA)
Stolte, E. (2018). Moving motivation: Increasing physical activity in people aged 50 and over.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
E-mail address:
vuresearchportal.ub@vu.nl
5
Process evaluation of a
internet-based intervention to
motivate adults aged 50+ to
engage in physical activity
The use of the Internet for interventions in the health prevention domain is a developing area (Griffiths, Lindenmeyer, Powell, Lowe, & Thorogood; 2006). Evaluating Internet interventions is im portant to improve knowledge among researchers and health professionals on the usefulness of inter vention components (Steele, Mummery,
& Dwyer, 2007). More specifically, there is a need to document and evaluate the components of complex interventions in order to increase the knowledge in the field of lifestyle interventions and develop effective interventions (Aalbers, Baars, & Olde Rikkert, 2011). Process evaluation follows a formative evaluation of acceptability and feasibility and provides a detailed evalua tion that supplements the effect evaluation provided by randomized controlled trials (Moore et al., 2015). In this chapter, we conduct a process evaluation of an online intervention program that was devel
oped to motivate adults aged 50 and older to be more physically active. The sixweek intervention program had four components: an online (1) PA plan and (2) diary, (3) PA tips, and (4) two fullcolor in formational brochures about physical activity. In addition, to study the effect of prompting on adher ence, email prompts were sent to half of the participants (prompting group), who were randomly as signed. See Chapter 3 for the full description of the intervention development, Chapter 4 for the evalua tion of feasibility and acceptability, and Chapter 6 for the effect study on physical activity and motiva tion. There are no standards for the content of a process evaluation, but in general, it examines the con text, implementation and mechanisms of impact (Moore et al., 2015). Process evaluation of the context examines how external factors influence the delivery and functioning of interventions. Process evalua tion of implementation involves examination of the structures, resources and processes through which delivery is achieved and the quantity and quality of what is delivered.
Mechanisms of impact refer to how intervention components and the interaction of participants with those components trigger change. The process evaluation of the current online intervention evaluates the implementation and the mechanisms of impact to derive recommendations for researchers and health professionals. The re
search questions on implementation and mechanisms of impact are discussed in more detail below. The context of the intervention is beyond the scope of our research. To evaluate contextual factors (for in stance, widespread attention for a specific health problem or measures at the national level to protect health), a comparison must be made with either the same context at another point in time or another context that is comparable in other factors. We did not measure such contextual factors.
We evaluate implementation by assessing which participants were reached in the light of the re cruitment efforts. Research question 1 asks to what extent the recruitment succeeded in reaching par ticipants with the characteristics we were aiming for.
Regarding participant characteristics, the target group of our program was people
5
aged over 50 living in the Netherlands with a need and a desire to be more physically active. Therefore, we needed to reach people who were not active enough and had the intention and ability to change their behavior but needed help to achieve their aim.
To reach the target group, we informed potential participants about the criteria and checked after finishing the study whether our criteria were met. We did not use exclusion criteria for activity level as we did not know in advance which type of people would respond to the program and which combination of criteria would be suitable to include the targeted group without excluding too many participants who might benefit from the program. However, a program should ideally include only those participants who can poten tially benefit from it in order to prevent disappointment among participants and avoid wasting the re sources of researchers and health professionals. Therefore, we assess whether inclusion criteria could have been beneficial and, if so, which criteria can be used in future programs. We do so by comparing the characteristics of the group that participated to those of the target group.
To assess the mechanism of impact, we evaluate how participants interacted with the interven tion components and how these components triggered change in short
term intervention goals. We perform this evaluation by assessing appreciation among participants. Appreciation is participants’ sub jective evaluation of what aspects they found pleasant and useful. Professionals who plan interventions need to know what type of participants are expected to appreciate a program. Research question 2 is:
To what extent do participants appreciate the intervention and its components? In answering research question 2, we also assess 2b: Which participant characteristics predict appreciation for the interven tion? To evaluate how the intervention components triggered change, we need to look at their impact on shortterm goals, which in turn are supposed to influence longerterm goals such as motivation and activity levels. The four shortterm goals are that older adults 1. know how much activity is necessary to meet the Dutch PA recommendations, 2. know which activities they can integrate into their daily life in order to become more active, 3. monitor their PA, and 4. set goals and make specific plans to be more physically active. We evaluate whether the intervention program is successful in increasing the four short
term goals. Research question 3 is: To what extent are the shortterm goals of the intervention reached?
Method
The evaluation was performed with information from two stages of research, i.e., a pilot stage and a main trial for establishing effectiveness. Before the intervention started, background characteris tics and baseline values of outcomes targeted by the program were collected with an online question naire. During the program, several outcomes were monitored with weekly questionnaires. At the end of the sixweek program, an evaluation of the program and its components was part of the questionnaire. More details on the study design and procedure are reported in Chapter 6.
Intervention content
The intervention consisted of an online PA plan and diary, online PA tips and two fullcolor infor mational brochures about physical activity, one from the Dutch Heart Foundation (2012) and one from an elderly foundation (Unie KBO & NISB Consult, 2012) (see Chapter 3). After the pilot phase, small changes were made to the program based on user feedback (see Chapter 4). PA tips in the pilot phase were either videos containing animated characters or text with the same content. In the main trial, the videos and text description were combined. After the pilot, a feedback element was also added to the PA diary in the form of a traffic light (green, orange, red), and the total amount of PA per day was added to the diary. Additionally, for the main trial, the general information on the website targeting potential participants was adapted to make it clearer that the program was aimed at people who wanted to be come more active. The information also stated that the program did not include personal coaching and that the objective was for participants to become their own coach.
Recruitment of participants
For participant recruitment, various methods were used at different stages of the program devel opment. For the pilot, participants were recruited using an advertisement on an online social platform for people over 50 and through online ads using Google AdWords. These advertisements were visible for only a limited time, and the response on the online platform was not high enough to use it as an outlet for the main trial. For the main trial, a press release was used to advertise the study.
The online adver tisement was still available for websites that were willing to place a written item about the program. The advertisement was placed on several online health platforms and in four email newsletters, of which two were sent by the Vrije Universiteit Amsterdam and aimed at participants in elderly education and at alumni.
One national newspaper wrote an article about the program. In total, 251 people signed up, of whom 223 also followed through with the registration and started the
5
program. The evaluation ques tionnaire at the end of the sixweek program was filled out by 161 participants.
Measurements Reach
The reach of the intervention was evaluated by means of questions from the registration ques tionnaire. Recruitment strategies were evaluated by asking participants how they found out about the program. The type of participant was evaluated by collecting data on the characteristics of participants who registered and participated. In the first questionnaire, before the start of the intervention program, participants’ date of birth, gender, educational level, height and weight were assessed.
Age was calcu lated by collecting their date of birth. Educational level was assessed in categories that are relevant for the Dutch educational system and categorized into low, medium and high educational levels. Body mass index (BMI, kg/m2) was calculated from height and weight and further categorized following interna tional guidelines (World Health Organization [WHO], 2000). The revised Physical Activity Readiness Ques tionnaire (rPARQ) was used to evaluate whether the participant had contraindications for increas ing PA. If a participant answered yes to any of the questions on the rPARQ, he or she was asked to eval uate whether participating would be safe for him or her and first contact a physician if necessary (Cardi nal, Esters, &
Cardinal, 1996). General health was measured by means of the SF12 Health survey (Ware, Kosinski, & Keller, 1996), which has a mental subscale and a physical health subscale, each consisting of six items. Scales were constructed using formal scoring instructions (Ware, Kosinski, & Keller, 1996). Physical activity was selfreported with the physical activity scale for the elderly (PASE) (Harada, Chiu, King, & Stewart 2001;
Washburn, Smith, Jette, & Janney, 1993). It assesses total physical activity during the past week, including sports as well as recreational activities, household work and physical activity during (volunteer) work. Motivation for physical activity is measured by the sum of three items (a = 0.91), two related to effort and one to intention, measured on sevenpoint Likert scales. In questions with an open answer format, participants were asked what their reason was for signing up for the pro gram. Stage of change was assessed with one question (adapted from Marcus et al., 1992) that deter mined which of five stages described the physical activity stage the best: 1. ‘I am currently not regularly physically active and do not intend to change this’, 2. ‘I am currently not regularly physically active and am considering changing that’, 3. ‘I am already physically active but not sufficiently so and want to do more’, 4. ‘I am currently regularly physically active and started being active in the last six months’, and 5. ‘I am currently regularly physically active and have been so for more than six months’.
Appreciation
Several questions were asked in the evaluation questionnaire. We created a measure of overall ‘appreciation’ by combining ‘satisfaction’ and ‘activation’ scores (a = .68 .77).
Participants were asked to express their satisfaction with the overall program and each of the four components (1 = ‘very bad’ to 10 = ‘very good’). Participants were asked how much each part of the program helped them be more physically active (1 = ‘not at all’, 2
= ‘not really’, 3 = ‘a little’, 4 = ‘somewhat’, to 5 = ‘very much’). The combined appreciation score was computed by taking the average of the satisfaction score and twice the activation score for each component (range 110). The three online components (PA plan, diary and tips) were evaluated in more detail on their usefulness and appearance, with six (for plan) and seven (for diary and tips) questions rated on a scale from 1 =
‘totally disagree’ to 7 = ‘totally agree’. Additionally, in the main trial, participants could comment on the appeal of the videos of the PA tips in an openanswer format of which the content is summarized. Participants were asked which benefits they experienced from the program with multiple answers possible from a list of preset options (Table 2) and an openanswer category. Finally, participants were asked whether they would recommend the program to oth ers with the options to answer no, maybe, or yes.
Short-term program goals
For the first goal, knowledge of PA norms, we asked a multiplechoice question testing the knowledge of the content of the norms. The question had five options, and the correct answer was ‘at least 30 minutes of moderate intensity activity, such as walking or cycling at a moderate pace, at least five times a week but preferably more’. For the second goal, knowledge of suitable activities, we asked four questions with a scale from 1 ‘don’t agree at all’ to 7 ‘totally agree’. An example is ‘I know which activities/sports suit me well’. We computed the sum of the answers (a = 0.84). The third goal, moni toring of PA, was assessed by the number of weeks a diary was filled out on the website. The fourth goal, setting PA goals and making specific plans to be more physically active (goal setting), was assessed by two instruments: whether goals were set in the online PA plan, and the sum of two items (rated from 1 to 5;
never, rarely, sometimes, often, very often) that form the goal setting subscale of the physical activity selfregulation scale (PASR12; Umstattd, Motl, Wilcox, Saunders, &
Watford, 2009). The fifth goal, planning, was measured by use of the online planning tool. Furthermore, an implementation inten tion scale adapted from other studies was used (Rhodes, Blanchard, Matheson, & Coble, 2006; Rise, Thompson, & Verplanken, 2003) and assessed in five weekly questionnaires during the intervention period (a = 0.92 0.95). Four questions on when, where, what and how participants had planned physi cal activity were asked. The answering categories were from ‘1 no plans’
to 7 ‘detailed plans’. The sum of the answers was computed.
5
Statistical analyses
We used descriptive statistics to answer the first research question, about the reach of the pro gram and participant characteristics, and question 2a, about participants’
appreciation for the program and its components. Multiple regression was used for research question 2b to assess whether partici pant characteristics predict appreciation for the intervention program. The predictor variables were baseline scores for PA and motivation to become more active, BMI, age, physical and mental health, and dummy variables for educational level, intervention group, and gender. Appreciation for the total pro gram and each component were the dependent variables. For question 3, on change in shortterm goals, a dependent ttest was used to assess change in the knowledge of suitable activities, and a Chisquare test was used to assess the change in the knowledge of the PA norms. Multilevel analyses were used to assess change over time in shortterm goals (question 3) for measures that were assessed in more than two questionnaires. Implementation intentions were assessed during the intervention period in five weekly questionnaires, and goal setting was assessed before, during and after the intervention in seven questionnaires.
Results
Reach
Table 1 shows the results of the recruitment for the pilot and main trial by the type of recruitment channel. A total of 33 participants were recruited during the pilot of the program and 190 during the main trial of the program. The majority of pilot participants were recruited through an online social platform for people over 50.
For the main trial, participants were recruited through a variety of chan nels. The most common ways participants knew about the program were word of mouth, the newsletter for elderly education and the online social platform for people over 50.
Regarding the characteristics of participants who were reached, their age ranged from 50 to 93 (M = 62.5; Table 2). Most (n = 182) were between 50 and 69 years old. The majority of participants were female (73%). The majority were highly educated (69%);
almost two in ten had a medium level of educa tion (18%), and few had a low level of education (13%). The majority (61%) had no contraindication for increasing physical activity; 88 participants mentioned one or more contraindications, for a total of 126 contraindications. Often mentioned was the use of medicine for high blood pressure or a heart condi tion (59 participants). The second most common contraindication was having a bone or joint problem such as arthritis (29 participants). One person postponed study enrolment to ask a doctor for advice on participating; all others with a contraindication agreed that their health was not a problem when they wanted to be more physically active, and they signed the informed consent.
The average score on general health (SF12; Table 2) was similar to the scores of a Dutch group in a validation study (Gandek et al., 1998); the norm scores for the physical health scale are 47.9 (age 4564) and 43.3 (age 6574) and for the mental health scale are 51.4 (age 4564) and 53.1 (age 6574). The majority of participants evaluated their overall health as good, very good or excellent (Table 3); 27% experienced some limitations when performing moderate physical activity, and 37% experienced some limitations when climbing stairs. The average weight of participants fell into the category of overweight (World Health Organization [WHO], 2000), with a BMI just over the cutoff of 25 (M = 26.5); 39% had a healthy weight (Table 4), and four participants were morbidly obese.
5
Table 1. Channels through which participants were reached
Pilot Main Trial Total
Word of mouth 4 36 40
Newsletter elderly education 0 39 39
Online social platform 27 9 36
Newspaper article 0 24 24
Physiotherapy website/newsletter 0 23 23
Google 1 19 20
University alumni/employees 0 11 11
Elderly organization 1 4 5
Other 0 25 25
Total 33 190 223
Table 2. Characteristics at registration (N = 223)
M SD
Age 62.5 7.3
BMI 26.5 4.4
SF12-Physical 49.1 7.8
SF12-Mental 50.5 9.6
PASE 131.6 74.2
Motivation 5.3 1.6
Table 3. SF12 overall health item (N = 223)
n %
Excellent 9 4
Very good 52 23
Good 122 55
Moderate 38 17
Poor 2 <1
Table 4. BMI in categories (N = 223)
n %
Underweight (<18.5) 1 <1
Healthy weight (18.5-25) 86 39
Overweight (25-30) 99 44
Obese (30-40) 33 15
Morbidly obese (>40) 4 2
Most participants either were considering becoming more physically active (contemplation phase, 27%; Table 5) or were already physically active but wanted to do more (preparation phase, 46%). Fifty participants (22%) were already regularly physically active for more than six months (maintenance phase). The average physical activity score on the PASE questionnaire upon registering for the program was 131.6 (Table 2), which is similar to the mean score of 131.3 found in a US sample of inactive older adults (mean age 66) (Washburn, McAuley, Katula, Mihalko, & Boileau, 1999). Physical activity at registra tion was higher for participants in a higher stage of change (Table 5). Motivation was high at registration with a scale mean of 5.3 (range 17; Table 2). Participants’ motivation was on average high est for participants in the contemplation and preparation phases and lowest for participants in the maintenance phase.
An open question on the motivation to participate in the program showed that about half of the participants in the maintenance phase had reasons for participating other than wanting to become more active. Twelve mentioned they wanted to check whether they were sufficiently active. Fourteen said that they found it an interesting topic or an important study, or they were just curious of the program or the results of the study. Eleven participants mentioned wanting to be motivated to stay healthy or hav ing trouble keeping as active as they used to be because of circumstances.
Others mentioned reasons such as missing structure despite being sufficiently active, wanting extra information about PA, or find ing PA to be an important topic.
Appreciation
Our second question is to what extent participants appreciated the intervention and its compo nents (2a) and which participant characteristics influenced appreciation (2b). We evaluated participants’ opinion on the program and its components. All program components were overall rated as sufficient (average higher than 5.5).
The total program and the PA diary received the highest average grade, and the PA tips and plan received the lowest average grades (Table 6). A multiple regression of appreciation for the total program showed that gender and baseline motivation to become more active predicted appreciation, while intervention group, age, baseline health, PA, BMI, and educational level did not (R2 = .15, F (9, 129) = 2.5, p = .01). Women (b = .18, p = .04) had lower scores on program appreciation, and participants with higher baseline motivation to become more active (b = .25, p = .004) had higher scores on appreciation. For the appreciation for the heart foundation brochure, age and physical health were significant predictors (R2 = .14, F (9, 114) = 2.1, p = .03). Older Participants (b = .23, p = .01) and participants lower physical health scores (b = .19, p = .046) appreciated the heart foundation brochure more. For the appreciation for
5
the second brochure, from the elderly foundation (R2 = .19, F (9, 115) = 2.9, p = .004), in addition to age (b = .30, p = .001) and physical health (b = .25, p = .006), baseline PA level (b = .19, p = .048) was a significant predictor. For the appreciation for the PA diary (R2 = .16, F (9,128) = 2.7, p = .005), age (b = .17, p = .04), baseline motivation (b = .29, p = .001) and BMI (b = .24, p = .008) were predictors. There was no effect on appreciation for the PA plan (R2 = .24, F (9,44) = 1.5, p = .17) or the PA tips (R2 = .12, F (9,117) = 1.7, p = .096).
Table 5. Stage of change and average level of PA and motivation per stage at registration (N = 223) PASE Motivation
n M SD M SD
1. ‘I am currently not regularly physically active and do not intend to change this’ 1 55.1 5.0
2. ‘I am currently not regularly physically active and am con sidering changing that’ 61 102.3 65.3 5.7 1.3 3. ‘I am already physically active but not sufficiently so and want to do more’ 103 128.2 59.3 5.7 1.0 4. ‘I am currently regularly physically active and started being active in the last six months’ 8 150.4 96.2 5.5 1.4 5. ‘I am currently regularly physically active and have been so for more than six months’ 50 173.1 89.5 3.8 1.9
Table 6. Appreciation (range 110) for the program (158 ≥ N ≥ 140)
M SD
PA plan 6.2 1.8
PA diary 7.0 1.7
PA tips 5.6 1.9
Brochure heart foundation 6.0 1.5
Brochure elderly foundation 6.2 1.6
Total program 7.1 1.6
Note. The number of answers for the PA plan is 73 due to a routing mistake in the questionnaire; not every participant who saw the PA plan graded it.
A detailed evaluation of the online components (Figure 1) shows that the PA diary was appreci ated most, with average scores higher than the middle of the scale on all items. The PA plan received a good evaluation on how it looked and on how easy it was to fill out, with an average score above the middle of the scale. The PA plan scored below the middle of the scale on items related to how good it made participants feel and whether it helped them achieve their goals. PA tips were not appreciated much, with all scores on average on the negative side of the scale. In an openanswer category, 45 peo ple commented on the PA tips: 27 of these comments
were negative, six were positive, and sixteen were mixed. Multiple participants mentioned that the information was too simple (n = 12). They also men tioned that the animated characters in the videos were too much like a cartoon and/or childish (n = 8) or that the voices sounded too computerized (n = 7). A few participants said they preferred textual infor mation, preferred videos with real people or preferred much more specific advice on special topics in stead of general information about PA.
When asked which benefits (Table 7) they experienced from the program, almost eight out of ten participants mentioned insight into PA patterns. The second most mentioned benefit was more disci pline for PA by almost half of the participants. New knowledge was mentioned by one in ten partici pants, and more confidence was mentioned by one in twenty participants. Of the 13 participants who mentioned a benefit in the open answer category, six said that they became aware of the importance of PA.
Most participants (57%; N = 157) said they would recommend the program to other people who want to become more active. About a third was not sure (33%), and one in ten would not recommend the program to other people (11%).
Short-term intervention goals
Research question 3 was to what extent the shortterm goals of the intervention were reached. Four shortterm goals were evaluated. Knowledge of PA recommendations was already high at premeasurement, as 83% of the participants correctly answered the multiplechoice question about the content of the norms. This percentage increased (N = 160 participants who answered this question twice) from pre
measurement (83%) to postmeasurement (89%) (Χ2(1) = 6.4, p = .01). The knowledge of suitable activities (range 428) increased from premeasurement (M = 20.5, SD = 5.4) to postmeasure ment (M = 22.8, SD = 4.4; t(159) = 5.6, p = .00). The third short
term goal was that participants would monitor their PA. Participants filled out, on average, 4.5 diaries, ranging from 0 to 9 (SD = 2.2). Of the 223 registered participants, 77% (172) registered their PA online for at least three weeks of the interven tion (the minimum number intended when creating the program). More than half of the 223 partici pants (53%, n = 117) filled out all six diaries during the program. The fourth shortterm goal was goal setting. The 121 participants who filled out the PA plan set in total 462 goals (an average of 3.8 goals per person; Table 8). The goals mentioned most were to improve or maintain health (identified by 98 par ticipants) and fitness (83 participants).
Figure 1. Average appreciation for online components of the intervention
5
Note. 1 = ‘totally disagree’ to 7 = ‘totally agree’, N = 158 for diary, N = 157 for plan and N = 133 for tips
Table 7. Benefits experienced by participants (N = 161)
%
Insight into my PA patterns 66
More discipline for PA 39
New ideas for PA activities 14
New knowledge on PA 9
Other benefits 8
More confidence regarding PA 4
None of the above 10
Table 8. Goals set by participants in the online PA plan (N = 121)
n
Improve or maintain health 98
Improve fitness 83
Lose weight 70
Achieve PA norms 69
Become more flexible 44
Gain strength 39
Improve balance 36
Other 23
Gave me a good feeling Reminded me of my intentions Helped me achieve my goals
PA tips PA plan PA diary
Helped me register my PA Videos were nice to watch Videos had added value
1 2 3 4 5 6 7
Useful Easy to fill out Looked nice
The goal setting scale did not present an increase (F(1,196) = 2.0, p = .16) over time in a multilevel analysis; the scores were M = 5.3 (SD = 2.1) before the start of the intervention program, M = 5.5 (SD = 1.9) during it and M = 5.5 (SD = 1.9) just after the program. The implementation intentions scale, which measured specific planning during the program, also did not change (F (1, 163) = 0.49, p = .48). Imple mentation intentions were M = 15.2 (SD = 7.5) in the first week and varied between M = 17.8 (SD = 6.6; week 2) and M =16.0 (SD = 7.1; week 4) in the other weeks of the intervention.
Discussion
In this chapter, we conducted the process evaluation of an Internetbased program for promoting PA in people aged 50 years and older. The aim was to inform researchers and health professionals about the recruitment and inclusion of participants (implementation) and the content (mechanisms of impact) of Internetbased physical activity (PA) programs for adults over 50. By means of online questionnaires, the reach, appreciation, and effectiveness in terms of achieving shortterm goals of the program were assessed, along with the influence of participant characteristics on appreciation and shortterm goals.
Recruitment
For research question 1 on recruitment, we examined to what extent the program reached par ticipants with the characteristics we were aiming for. In terms of finding study participants with a di verse age range, the recruitment was successful. Word of mouth was one of the most successful re cruitment channels. The use of multiple recruitment channels, including word of mouth, was sufficient for this study. The majority of participants were overweight or obese, which is a risk factor for chronic disease, and an increase in PA is beneficial for their health (American College of Sports Medicine [ACSM], 2009). Four participants were morbidly obese. Increasing activity is very challenging in a mor bidly obese condition due to strain on the joints. Therefore, in hindsight, it might have been better to refer them to an intervention that (also) focuses on diet and weight loss.
Some participant characteristics did not match the requirements of the target group.
Almost a quarter of participants did not belong to the target group: they were already physically active for more than six months, had an aboveaverage amount of PA, and they were poorly motivated to increase their PA. This group can be described as an active aging group that participated because they were interested in the topic or wanted to confirm that they were sufficiently active. The intervention recruited
5
partici pants who were highly educated. Although no specified target was established beforehand, we see this as unsatisfactory. It is unknown whether this was the result of the recruitment channels, as one suc cessful channel was a university newsletter.
It can also be that an Internetbased program attracts highly educated participants.
A review on Internetbased health interventions also found a selection bias of highly educated people (Aalbers, Baars, & Olde Rikkert, 2011). To avoid oversampling highly educated partici pants, it would be better to also specifically target lower educated participants in recruitment strategies. All studies included in the review by Aalbers, Baars, and Olde Rikkert (2011) used offline recruitment strate gies. Online recruitment has the potential to reach a large population, so more research is needed to examine effective online recruitment strategies in order to reach the right target group.
The strategy of communicating through text on the website about which people the intervention was suitable for was not successful at drawing the right target group.
Screening can be used to prevent participants who are already active from joining.
This screening can have two functions, i.e., informing people who are curious to know whether they are active enough (for which the screening would need to be quite thorough) and keeping participants who are already sufficiently active from joining the program and being disappointed. Additionally, offering a preview of the intervention content through a demo video could give participants a better idea of what to expect from the program.
Appreciation
For the mechanisms of impact, we asked: To what extent do participants appreciate the interven tion and its components? and 2b: Which participant characteristics predict appreciation for the inter vention? Overall intervention materials were appreciated with sufficient grades. The total program and the PA diary had the highest average appreciation scores. Higher appreciation for the program was as sociated with being male and having higher motivation levels at baseline. Appreciation for the brochures was associated with higher age and lower physical health scores. Appreciation for the PA diary was asso ciated with higher age and motivation and lower BMI. A more detailed evaluation of the online compo nents (diary, plan and tips) showed that the PA diary was appreciated well in all aspects. PA tips were appreciated least. According to the participants, the PA plan looked nice and was easy to fill in, but it did not make them feel good or help them achieve their goals. Of interest is that the average score on ap preciation for the total program was higher than the scores of its components.
This suggests that par ticipants based their total appreciation on the components they appreciated most and possibly also on the general opportunity to participate in this program. The benefits of the program, as experienced by participants, were
insight into PA patterns and more discipline for PA. New knowledge about PA and confidence in performing PA were rarely mentioned as benefits.
Short-term program goals
The third research question is to what extent the shortterm goals of the intervention were reached. Knowledge of the PA norms and knowledge of suitable PA activities were already high at the registration for the intervention and increased after the intervention. Another shortterm goal was for participants to monitor their PA, which was successfully achieved by the majority of participants using the online PA diary.
The last shortterm goal was for participants to set goals and make specific plans to be more physically active. A fairly large group of participants did not use the online PA plan. Those who used the plan set many types of goals. Overall, the effectiveness of the program to reach shortterm goals could be better.
Program content
PA diary was the most successful element of the intervention with regard to appreciation and reaching the shortterm goal of monitoring PA. Furthermore, the most mentioned benefit of the pro gram was insight into PA. An online PA diary or another form of registering behavior is therefore a rec ommended element for an online PA program for older adults.
The use of a PA planning schedule in this setting is challenging. Although participants on average thought the PA plan looked nice and was easy to fill in, it did not make them feel good or help them achieve their goals. One interpretation is that the PA plan made participants feel more aware of the goals that they did not achieve.
When applying an online planning instrument, it is important to create an engaging and enjoyable user experience. One possibility might be the use of gamification to increase the use and enjoyment of a planning instrument (OinasKukkonen &
Harjumaa, 2009).
The PA tips with animated characters were not highly appreciated. The low appreciation could be attributed to multiple things. The form of the videos might have been unappealing as the spoken text, which was created with texttospeech software, was not perfectly fluent for the Dutch language. Some participants mentioned they did not like the cartoon style of the videos or the computerized voices.
Animations are used in many current Internet applications, from customer service tools to elearning modules, but older adults may not be exposed to them as much as younger generations; therefore, they may associate them exclusively with childhood.
Furthermore, the content of the PA tips was partly com plementary to the brochures,
5
presenting the same information in a different way to make it stick. Par ticipants mentioned that the content was too simplistic. As this group was highly educated, the partici pants might have already known much of the information provided in the program. In sum, recommen dations based on participants’ experience with the PA tips in this program are that videos with real peo ple might be more attractive than animated characters, texttospeech software is not recommended, and information should be catered to the specific target group. The use of animated videos in this age group needs more study, and if applied, a usercentered design approach (Van GemertPijnen et al., 2011) may help make videos appealing to this specific group.
Limitations
The limitations of this study are that we did not reach participants with a lower educational sta tus, and therefore, we do not know whether the program would have been appreciated differently with more variation in participants’ educational background. Moreover, some participants did not fit the characteristics of the target group that we aimed for. Therefore, measures of appreciation may have been higher if participants who did not match the targeted group had been excluded from the program beforehand. Conclusions from this study need to be interpreted with care as this was a specific Internetbased program targeted at older adults. Some of the findings might be related to the specific group of participants or the specific way the components of the program were constructed.
Conclusion
We conclude that there is a demand for an Internetbased program to promote PA in older adults. The recruitment was successful in attracting participants, and the program overall was well appreciated, although there was variation in the appreciation for different components. Paperbased information and an online PA diary were appreciated by this age group. The online PA plan and animated videos were not appreciated much. The use of animated videos might not be suitable for this target group and should be studied more. The shortterm outcomes of increasing knowledge and registering PA were reached, while the use of goal setting and planning was infrequent. Improvements can be made in reaching the target group and on some components of the program. We recommend selecting potential participants beforehand and using specific recruitment strategies to target people with a low educa tional level. The use of Internetbased programs as part of a range of intervention instruments to target physical activity in older adults is a promising endeavor, and they need further development.
