Using Ecological Momentary Assessment to Evaluate Current Physical Activity
Objective. The purpose of this study was to assess the value of ecological momentary assessment in evaluating physical activity among children, adolescents, and adults. It also determines whether ecological momentary assessment fulfills the criteria of validity, reliability, objectivity, norms, and standardization applied to the tools used for the evaluation of physical activity. Methods. The EBSCO-CINHAL, Medline, PsycINFO, PubMed, and SPORTDiscuss databases were reviewed in December 2012 for articles associated with EMA. Results. Of the 20 articles examined, half (10) used electronic methods for data collection, although various methods were used, ranging from pen and paper to smartphone applications. Ten studies used objective monitoring equipment. Nineteen studies were performed over 4 days. While the validity of the EMA method was discussed in 18 studies, only four found it to be objective. In all cases, the EMA procedures were precisely documented and confirmed to be feasible. Conclusions. Ecological momentary assessment is a valid, reliable, and feasible approach to evaluate activity and sedentary behavior. Researchers should be aware that while ecological momentary assessment offers many benefits, it simultaneously imposes many limitations which should be considered when studying physical activity.
Physical activity (PA) is understood in many ways. Caspersen et al.  define PA as “any bodily movement produced by skeletal muscles which results in energy expenditure” [1, 2]. However, another definition, accepted by social scientists, states that PA is “behavior which can be conceptualized on a continuum from minimal to maximal movement” . The term “exercise,” defined as “a subset of PA that is planned, structured, and repetitive bodily movement done in order to improve or maintain one or more physical components of physical fitness,” is also frequently used  and accepted by the American College of Sports Medicine (ACSM).
In addition to PA, two other terms are often defined while setting reference standards or benchmarks for children and adolescents: “physical inactivity” and “sedentary behavior.” In children, physical inactivity is defined as a “PA level that is lower than in healthy individuals of similar age, gender, and cultural and socioeconomic background” . A sedentary lifestyle is described by the ACSM as “not participating in a regular exercise program or not meeting the minimal PA recommendations” . Marshall and Welk argue that an insufficient level of PA is not physical inactivity but “sedentary behavior” .
Physical activity (PA) can be assessed by objective and subjective methods, and these should fulfill several eligibility criteria: validity, reliability, objectivity, norms, and standardization [4, 5]. Objective methods are based on mechanical/electronic monitoring using accelerometers, pedometers, or heart rate monitors or on the measurement of such physiological parameters as direct and indirect calorimetry, maximum oxygen consumption/VO2max, doubly-labeled water consumption, or energy expenditure. Alternatively, subjective methods based on direct observation or questionnaires can be used [2, 3], and these can be applied to the assessment of sedentary behavior. The strength of the questionnaire method is its ability to assess the type, location, and prevalence of PA, for example, whether it occurs in school or during free time activities, with whom PA is undertaken, for example, with friends, family, or alone, and the emotions associated with PA. Furthermore, as objective methods do not differentiate between periods of inactivity and periods when the device is not being worn, subjective methods are preferred for measuring sedentary behavior . Within the group of subjective methods, the “daily life” sampling approaches, such as the experience sampling method (ESM) and ecological momentary assessment (EMA), have recently aroused considerable scientific attention . Some authors use the terms ESM and EMA interchangeably but Trull and Ebner-Priemer  emphasize that there are two different methods. EMA is viewed as a broader methodology that attempts to integrate a number of assessment traditions of self-reported subjective states and behavior, including ESM,  with the sampling and monitoring of physiological processes, behaviors, or states by electronic devices such as mobile phones .
In the fields of rehabilitation, sport science, and behavioral medicine, EMA has often been used to assess the dynamic changes in the behavior of children, young people, and adults, based on their PA and sedentary behavior . Not only does it allow data to be collected in real time , but also the clinical and theoretical complexity of PA can be examined in a real-world environment, along with its associated problems and dynamic processes . EMA analysis highlights individual differences in behavior, their distribution over time, the factors affecting their behavior, and the mutual associations between them. A comprehensive ecological approach provides rich, diverse, and complex data which can be used in the analysis of behavior .
Many studies have described the advantages of using EMA methodology to assess current activity [7, 10–12], the most important one being that the EMA approach combines both ecological and momentary aspects. Its ecological aspect is represented by data collection being performed in the real world, which allows the available information to be generalized to the real life of individuals, hence providing ecological validity. Its momentary aspect is defined by the real-time collection of information, in this case about PA, which minimizes errors caused by the necessity of recalling what had happened in the past. The difficulty of recalling the past often constitutes an obstacle to traditional methods of assessing PA, which usually restricts measurement to intervals of less than one day. However, as information is gathered at random intervals or is selected based on a pilot study during EMA analysis, it is possible to observe behavioral changes and the factors causing such changes, as subjects repeatedly answer the same set of questions during a single study [7, 10–12].
Data can be collected in several ways using the EMA approach, such as questionnaires, diaries, or electronic devices running special software. It can also be collected through “pencil and paper” questionnaires sent and verified each day during an ongoing study . Alternatively, electronic methods such as online questionnaires  or portable electronic devices such as PDAs (personal digital assistants) and mobile phones can be used [11, 13–15]. The PDA collects data, informs subjects when which they are supposed to complete the questionnaire, and sends the acquired data directly to the authors , while free, easy-to-use software such as MyExperience can be installed on mobile phones [16, 17].
The TelEMA platform, a telephone assessment platform for clinical and research applications, is regarded as an effective method of assessing behavior in real time by EMA. It sends signals and data between the experimenter and the test subject via the telephone network. The sent signal indicates a specific time to report current activity and integrates returning calls or text messages with the EMA survey. Data can be simultaneously collected this way by many researchers working in different locations on different projects .
The EMA approach allows data to be collected at regular intervals, such as once per day at a regular time or at random times during the day [7, 11]. If random requests are sent, the total number of samples must be first determined and must be the same for all individuals . The time of data collection may also be chosen depending on the aim of the study, for example, whether it addresses events or behavior which occur during the day or during PA .
The purpose of this study was to determine whether EMA is a suitable approach to evaluating physical activity among children, adolescents, and adults. A second aim is to determine whether EMA meets the criteria of measurement tools using to evaluate physical activity: validity, reliability, objectivity, norms, and standardization.
The EBSCO-CINHAL, Medline, PsycINFO, PubMed, and SPORTDiscuss databases were reviewed in December 2012. The inclusion criteria of reviews were that (1) the article had to be available in English, (2) the summary or title of the article had to include the following keywords “ecological momentary assessment” and “physical activity” or “ecologic momentary assessment” and “physical activity,” (3) EMA had to be used to evaluate PA and sedentary behavior, (4) the article had to be available as full text, and (5) the article had to be a research study. A summary flowchart of the search, selection, and inclusion process is presented in Figure 1.
A total of 53 unique articles were found in the EBSCO-CINHAL, Medline, PsycINFO, PubMed, and SPORTDiscuss databases. Thirty-two of these articles met the inclusion criteria in terms of language (1) and content of keywords in the title or summary (2). The keywords “ecological momentary assessment” and “physical activity” or “ecologic momentary assessment” and “physical activity” were found in the titles and summaries of 14 articles and the abstracts of 18 articles. Another 12 articles were rejected because they either did not use EMA to evaluate PA and sedentary behavior (3), were not available as full text (4), or were not research studies (5). In total, twenty articles were included in this review study [9, 17, 19–36]. The oldest article was published in 2005 , whereas the newest articles were published in 2011 [17, 27, 28, 36] and 2012 [29–31].
The included articles were divided into two groups. The first comprised 14 articles in which the subjects were children and adolescents (Tables 1 and 2) [9, 19–31]. The second group comprised 6 articles which focused on adults (Table 3) [17, 32–36]. Only one study concerned older people, aged 50 to 76 . Both girls and boys were used as subjects in 8 articles [20, 22–24, 27, 28, 30, 31], men and women in 3 [17, 34, 35], three only used girls [19, 21, 26], three only women [32, 33, 36], and three only boys [9, 25, 29]. No articles concerned men only.
The most common methods used to collect EMA data were traditional pencil and paper diaries [20–25, 32, 34, 35] and mobile phones with MyExperience software [17, 27–31]. However, diaries for handheld computers [9, 19, 33, 36] and telephone conversations were also used . Seven out of 10 studies were based on electronic methods of collecting data from children and adolescents [9, 19, 27–31].
Studies evaluating activity by EMA among children, adolescents, and adults were commonly conducted twice a year for 4 consecutive days [9, 19–22, 24, 25, 29, 30], over three extended weekends , or for several consecutive days [17, 23, 27, 28, 31–36]. In this last group, 50% of the studies evaluated PA at random times over 4 consecutive days: from Saturday to Tuesday-1 , from Friday to Monday-3 [27, 28, 31] and over 3 weekdays and 1 day over the weekend-1 .
Some studies supported the EMA approach by using such objective tools as accelerometers or heart rate monitors to assess activity [9, 17, 19, 26–31]. Another tool was the talk test, which evaluates the intensity of PA . Such additional tools were most commonly used to assess activity of children and adolescents [9, 19, 26–31].
Furthermore, all studies were evaluated with regard to the choice of activity evaluation tool and its measurement characteristics. While the validity of the EMA method was examined in 18 out 20 studies [9, 17, 19–32, 35, 36], the remaining two cases referred to a different study [33, 34]. The reliability of the EMA approach was not mentioned in only one study . The objectivity of the EMA approach was demonstrated in one study , whereas another three [19, 22, 26] referred to other articles for analysis. None of the remaining studies mentioned objectivity. Norms were not described in any of the included articles. In all cases, the EMA procedures were precisely documented and confirmed to be feasible (standardization).
4. Discussion and Conclusion
The first aim of this study was to investigate the validity of using ecological momentary assessment (EMA) to evaluate physical activity (PA) among children, adolescents, and adults. The review compiled articles ranging from 2005 to 2012. The relatively short time-span and small number of publications related to this problem suggest that research into the use of EMA to evaluate PA among children, adolescents, and adults is at an early stage.
While 14 articles concerned PA and sedentary behavior among children and adolescents, another 6 addressed it use on adults, and only one publication evaluated activity among the elderly. Moreover, data collected electronically as telephone surveys and electronic diaries predominated among studies focused on children and adolescents. The greater prevalence of electronic forms of collecting EMA data among studies addressed at young people could be due to its easiness of use and the lack of resistance associated with the use of modern technology demonstrated by this specific population. Furthermore, keeping electronic diaries or completing EMA questionnaires on handheld computers or phones could be more attractive for young people than the traditional pencil and paper form of data collection. Older people have been found to be reluctant to participate in studies which promoted electronic forms of data collection . However, it should be noted that information and communications technology is becoming an increasingly common mode of data collection in medicine and health promotion, the main reasons being the lower costs of research and treatment and the more widespread availability of such technology [13, 14].
Many studies use objective measurement tools, such as an accelerometer or a heart rate monitor alongside the subjective EMA assessment. By comparing the data recorded using these devices with the EMA data, it was possible to determine the differences and compliances between the two sources and confirm the construct validity of the study [9, 17, 26–31]. Moreover, most studies using objective measurement tools were intended to evaluate the PA among children and adolescents. Hence, most authors identified whether children were experiencing difficulties and wanted to complete EMA surveys at the time they were involved in some form of activity.
Data comparison provided an indication of the intensity of PA undertaken at the time of submitting the EMA surveys. However, some studies note no significant correlation between the number of steps on a pedometer, or moderate-to-vigorous physical activity (MVPA), and the number of survey questions which remained unanswered . This might indicate a lack of precision in determining a connection between the level of activity intensity, the responses of the subject, and time that the signaled questionnaires were submitted. On the other hand, there might simply be no relationship between the MVPA value and the survey responses. The combination of activity assessment using subjective EMA tools and the evaluation of depression, pain, environmental assessment, or the level of activity intensity using other scales or tests allow accompanying moods, stress levels, pain, barriers to undertaking PA, and levels of PA intensity to be explored [34, 36, 38–40].
An interesting aspect of the evaluated studies was the design and duration of their tests. Many studies [9, 19–25, 29, 30] evaluated the activity twice at given intervals, that is, twice a year, usually in early fall and in the spring. Such a methodology, used in studies focused on children and younger subjects, indicates that the authors intended to demonstrate seasonal variation, changes in PA by the study group depending on the time of year. Moreover, the selection of consecutive days for testing, usually encompassing both weekdays and weekends, allowed observers to evaluate weekly fluctuations in PA participation.
The second aim of this study was to analyze the employed EMA tools in terms of their measurement characteristics. The findings demonstrate that the EMA approach constitutes a valid, reliable, and feasible measurement tool, which clearly indicates that EMA can be considered a suitable method for assessing PA among children, adolescents, and adults.
One of the benefits of using EMA to assess PA among groups of children, adolescents, and adults is that it allows the researcher to gain an insight into factors associated with the activity or behavior itself. Heron and Smyth (2010)  emphasize the benefits of the EMA approach. They note that as data collection takes place either at a specific time or during an activity of interest to researchers, the measurement is free from the defects associated with the need for the respondent to remember past events. A second, equally important value of the EMA approach is its ecological validity, that is, that its results can be generalized by its ability to perform measurements in the real world: the authentic surroundings of the respondents. Another advantage is its ability to measure several factors at the same time, for example, PA and the accompanying mood , which provides researchers with an opportunity to conduct detailed and more comprehensive examinations, such as identifying relationships between common activities and behavior. These advantages of using EMA to assess PA and sedentary behavior were noted in all studies analyzed in the course of this paper.
There are, however, some limitations associated with EMA and the studies which use it to evaluate PA. A major difficulty is associated with the respondent providing answers in pencil and paper form while involved in PA, particularly when it is intense [9, 19, 30, 31], and more generally, EMA is considered a more burdening and time-consuming approach for participants in comparison with retrospective methods, despite the potential benefits of real-time activity examination . Moreover, adolescents were often found to be reluctant to complete a signaled EMA survey, due to reasons such as a chaotic environment during the survey or the presence of psychological or behavioral problems [9, 19]. Another limitation reported by researchers is the inability to compare EMA results with metabolic energy expenditure, such as kilocalories or metabolic equivalent [9, 20–25, 32–35]. In addition, some studies used EMA only to determine the frequency and duration of the activity, which was considered by some researchers as a limitation of daily PA assessment [19, 28, 30, 31]. Finally, one weakness possessed by all analyzed studies was either an imprecise definition of PA or its absence; in some cases, such activities as active transport to and from school, or activities during school hours, were not taken into account while assessing PA [28, 31].
Furthermore, key similarities and differences exist between the traditional and EMA approaches to assess PA. With regard to the similarities, it is possible to use the same forms of data collection, such as questionnaires or diaries, in both cases [2, 11, 41], and in both cases, subjects may make mistakes while submitting their answers through study protocols [9, 19, 30, 31, 41]. Both EMA and the traditional approach can also be burdening and demanding for the respondents in terms of the time and effort needed to provide required information [25, 41]. However, the obvious benefit of EMA is its ability to collect data in real-time and real-world circumstances [7, 10–12]. The traditional approach to assess PA requires the subjects to recall specific things that occurred at specific times in the past, which may result in inconsistencies and errors in provided information [7, 10–12, 41].
Finally, it should be noted that EMA is coherent with the ecological task analysis (ETA) described by Davis and Burton (1991)  and is widespread in the area of adapted physical activity (APA). It has been mentioned that EMA allows the PA profiles of children, adolescents, and adults to be assessed in real-world circumstances and either at specific or at recurring points in time [7, 10–12]. As ETA is a model which considers complex relationships between the individual, the environment, and the task and is used as a system of evaluation and instruction , a combination of EMA and ETA may allow the factors inducing engagement in PA to be determined, as well as the influence of environmental limitations on participation in PA. In addition, an ecological approach, taken by both EMA and ETA, allows affective mechanisms and cognitive abilities, or psychomotor skills that have a significant impact on engaging in PA, to be analyzed.
The use of EMA, in its various forms, to assess PA and sedentary behavior is a valid, reliable, and feasible method of evaluation. However, in the assessment of PA, it is important to define what “physical activity” actually is. Researchers should also be aware that apart from significant benefits offered by EMA, there are various limitations to this approach which should be taken into consideration when attempting to conduct observations of PA.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
C. J. Caspersen, K. E. Powell, and G. M. Christenson, “Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research,” Public Health Reports, vol. 100, no. 2, pp. 126–131, 1985.View at: Google Scholar
W. Osiński, “Assessment of the level of physical activity,” in Anthropomotoric, W. Osiński, Ed., pp. 320–330, AWF Poznań, 2003.View at: Google Scholar
S. J. Marshall and G. J. Welk, “Definitions and measurement,” in Youth Physical Activity and Sedentary Behavior: Challenges and Solutions, A. L. Smith and S. J. H. Biddle, Eds., pp. 3–29, Human Kinetics, 2008.View at: Google Scholar
B. N. Strand and R. Wilson, “Selecting and constructing tests,” in Assessing Sport Skills, B. N. Strand and R. Wilson, Eds., pp. 9–22, Human Kinetics, 1993.View at: Google Scholar
W. Osiński, “Eligibility criteria for tests in studies of physical fitness and motor skills,” in Anthropomotoric, W. Osiński, Ed., pp. 154–166, AWF Poznań, 2003.View at: Google Scholar
T. J. Trull and U. W. Ebner-Priemer, “Using Experience sampling methods/ecological momentary assessment (ESM/EMA) in clinical assessment and clinical research: introduction to the special section,” Psychological Assessment, vol. 21, no. 4, pp. 457–462, 2009.View at: Google Scholar
A. A. Stone and S. Shiffman, “Ecological momentary assessment in behavioral medicine,” Annals of Behavioral Medicine, vol. 16, pp. 199–202, 1994.View at: Google Scholar
D. S. Moskowitz and S. N. Young, “Ecological momentary assessment: what is it and why it is a method of the future in clinical psychopharmacology,” Journal of Psychiatry and Neuroscience, vol. 31, no. 1, pp. 13–20, 2006.View at: Google Scholar
J. Froehlich, J. Landay, M. Chen, S. Consolvo, B. Harrison, and I. Smith, “An overview of in situ self report and the my experience tool,” 2006, http://www.learningace.com/doc/588191/0c5f44c6c4d850d0becb1b62cef01cc4/selfreportandmyexperience2006.View at: Google Scholar
G. F. Dunton, C. K. Whalen, L. D. Jamner, and J. N. Floro, “Mapping the social and physical contexts of physical activity across adolescence using ecological momentary assessment,” Annals of Behavioral Medicine, vol. 34, no. 2, pp. 144–153, 2007.View at: Google Scholar
T. Gorely, S. J. Marshall, S. J. H. Biddle, and N. Cameron, “The prevalence of leisure time sedentary behaviour and physical activity in adolescent girls: an ecological momentary assessment approach,” International Journal of Pediatric Obesity, vol. 2, no. 4, pp. 227–234, 2007.View at: Publisher Site | Google Scholar
S. J. H. Biddle, T. Gorely, S. J. Marshall, and N. Cameron, “The prevalence of sedentary behavior and physical activity in leisure time: a study of Scottish adolescents using ecological momentary assessment,” Preventive Medicine, vol. 48, no. 2, pp. 151–155, 2009.View at: Publisher Site | Google Scholar
T. Gorely, A. J. Atkin, S. J. H. Biddle, and S. J. Marshall, “Family circumstance, sedentary behaviour and physical activity in adolescents living in England: Project STIL,” International Journal of Behavioral Nutrition and Physical Activity, vol. 11, no. 6, article 33, 2009.View at: Publisher Site | Google Scholar
G. F. Dunton, L. Yue, S. S. Intille, J. Wolch, and M. A. Pentz, “Physical and social contextual influences on children's leisure-time physical activity: an ecological momentary assessment study,” Journal of Physical Activity and Health, vol. 1, no. 8, pp. 103–108, 2011.View at: Google Scholar
G. F. Dunton, K. Kawabata, S. S. Intille, J. Wolch, and M. A. Pentz, “Assessing the social and physical contexts of children's leisure-time physical activity: an ecological momentary assessment study,” The American Journal of Health Promotion, vol. 26, no. 3, pp. 135–142, 2012.View at: Google Scholar
G. F. Dunton, A. A. Atienza, C. M. Castro, and A. C. King, “Using ecological momentary assessment to examine antecedents and correlates of physical activity bouts in adults age 50+ years: a pilot study,” Annals of Behavioral Medicine, vol. 38, no. 3, pp. 249–255, 2009.View at: Publisher Site | Google Scholar
M. Kanning and W. J. Schlicht, “Be active and become happy: an ecological momentary assessment of physical activity and mood,” Journal of Sport & Exercise Psychology, vol. 32, no. 2, pp. 253–261, 2010.View at: Google Scholar
D. S. Bond, J. G. Thomas, B. A. Ryder, S. Vithiananthan, D. Pohl, and R. R. Wing, “Ecological momentary assessment of the relationship between intention and physical activity behavior in bariatric surgery patients,” International Journal of Behavioral Medicine, vol. 20, no. 1, pp. 82–87, 2011.View at: Publisher Site | Google Scholar
D. E. Conroy, S. Elavsky, A. L. Hyde, and S. E. Doerksen, “The dynamic nature of physical activity intentions: a within-person perspective on intention-behavior coupling,” Journal of Sport & Exercise Psychology, vol. 33, no. 6, pp. 807–827, 2011.View at: Google Scholar
J. R. Sirard and R. R. Pate, “Physical activity assessment in children and adolescents,” Sports Medicine, vol. 31, no. 6, pp. 439–454, 2001.View at: Google Scholar
W. E. Davis and A. W. Burton, “Ecological task analysis: translating movement behavior theory into practice,” The Journal of Physical Chemistry A, vol. 8, pp. 154–177, 1991.View at: Google Scholar
M. Bouffard and W. B. Stream, “Critical thinking and professional preparation,” in Adapted Physical Activity, R. D. Steadward, G. D. Wheeler, and E. J. Watkinson, Eds., pp. 1–10, University of Alberta Press, Edmonton, Canada, 2003.View at: Google Scholar