Journal Logo

Critical Measurement Issues Related to Walking

Walking and Measurement

BASSETT, DAVID R. JR.1; MAHAR, MATTHEW T.2; ROWE, DAVID A.3; MORROW, JAMES R. JR.4

Author Information
Medicine & Science in Sports & Exercise: July 2008 - Volume 40 - Issue 7 - p S529-S536
doi: 10.1249/MSS.0b013e31817c699c
  • Free

Abstract

Walking is a fundamental form of human locomotion. It is practiced worldwide by people of all cultures and races and by nearly all age groups (except for the very young and very old). Walking is an integral part of many tasks of daily life, and it is one of the most common forms of leisure-time physical activity (LTPA) (32). As a group, walkers are healthier than their sedentary counterparts, and they have lower blood pressure and more favorable body mass index values. From prospective longitudinal studies, we know that walking is associated with a reduced risk of coronary heart disease (13,17,19).

Exercise scientists are interested in measurement of walking for many reasons. Accurate valid measures of walking are needed to compare the physical activity behaviors of various populations. They also allow dose-response relationships between walking and health variables to be measured. Valid measures of walking enable us to track time trends in this vital physical activity and to measure the impacts of public health efforts to increase walking. Key aspects of walking behaviors of interest to researchers are walking time, duration, speed, purpose, leisure, transportation, and work activities. Measuring each of these areas provides insight to the relations between walking behavior and health.

WALKING ASSESSMENT TOOLS

Measurement of walking is not new. In ancient Rome, distances were measured by counting steps. In fact, the word "mile" is derived from the Latin phrase, milia passuum, meaning 1000 paces. The Roman mile was 1000 paces (or 2000 steps) of an adult citizen.

In this section, we will consider a variety of instruments used to measure walking behaviors, including pedometers, accelerometers, trail counters, direct observation, and physical activity questionnaires. With these assessment tools, scientists can measure walking at the personal level, community level, and national or international level.

Pedometers.

Pedometers are small devices, worn on the body, that measure steps or distance. Five hundred years ago, Leonardo da Vinci designed an early pedometer (6). This device, worn at the waist, had a protruding vertical lever arm that was affixed to the thigh. When the individual walked, the movement of the lever arm caused a gear-and-ratchet mechanism to rotate and steps to be recorded.

In the 1990s, electronic pedometers were developed that use a spring-suspended lever arm that moves up and down as a person walks. With each step, an electrical contact is made and a step is recorded. The accuracy of electronic pedometers has been studied, and the best models were found to measure steps to within 3% of the actual values, at walking speeds of 3 mph and faster (3,9,28). However, most pedometers undercount steps at slow walking speeds, by various amounts. Some pedometers allow stride length to be entered and provide distance estimates that are reasonably valid at normal walking speeds, but they overestimate distance at slow speeds and underestimate it at fast speeds due to variations in stride length. In addition, some pedometers allow body weight to be entered and provide calorie expenditure estimates. Although these calorie estimates are reasonably accurate (±20%) for walking at speeds of 2 to 4 mph, they vastly underestimate the energy cost of most other activities (4). The Yamax SW200 digiwalker, New Lifestyles NL2000, and Walk 4 Life LS2525 are some of the commonly used pedometers in research.

Accelerometers.

Accelerometers are small devices, usually worn on the belt, which measure instantaneous accelerations, for the purpose of estimating energy expenditure or time spent in various intensity categories. Traditionally, these contained a horizontal cantilevered beam with a weight on it; when exposed to vertical acceleration, the beam flexes and compresses a piezoelectric crystal that generates a voltage proportional to the acceleration. The newer integrated chip sensors have a seismic mass that sits directly over a piezoelectric element (5). The raw acceleration signal is then put through an analog-to-digital converter, filtered, and rectified. Using an internal clock, the acceleration units are accumulated over a discrete, user-specified epoch. Most accelerometers now have enough internal memory to store minute-by-minute acceleration data for several weeks. The Actigraph, Actical, Caltrac, and BioTrainer are some of the commercially available accelerometers used in research.

Montoye et al. (21) conducted an early study on the relationship between accelerometer counts per minute and energy expenditure. While wearing a uniaxial accelerometer on the back at the waist, subjects performed 14 activities including bench stepping, knee bends, floor touches, and treadmill walking or running at various speeds and grades. The correlation between accelerometer readings and oxygen uptake V˙O2 (mL·kg−1·min−1) was r = 0.74. One disadvantage of the Caltrac was that it tended to overestimate the cost of walking at 3-4 mph by 20-50% (2,14).

Freedson et al. (10) performed a metabolic calibration of the CSA Actigraph, a uniaxial accelerometer. They had subjects walk at 3 and 4 mph on the treadmill and jog at 6 mph. Using this restricted group of activities, the correlation between accelerometer counts per minute and energy expenditure (METs) was r = 0.88. The algorithm that was generated was later shown to have reasonable accuracy for predicting the energy cost of level walking, but it underestimated the cost of other light-to-moderate lifestyle activities, including household chores, gardening, and mowing, as well as recreational activities (e.g., tennis, golf) (4).

The International Life Sciences Institute funded a series of studies on the ability of accelerometers to measure moderate-intensity lifestyle activities (15,30,33). Hendelman et al. (15) and Swartz et al. (30) generated new CSA regressions based on a wide range of activities; although these equations were shown to be more successful at predicting the energy cost of moderate-intensity lifestyle activities than those based on walking and jogging (e.g., Freedson's equation), they were later shown to overpredict sedentary behaviors, light activities, and walking. A promising new method of analyzing accelerometer data involves distinguishing between rhythmic locomotor activities such as walking and jogging and other types of intermittent activities based on a) the mean accelerometer counts and b) the variability in accelerometer counts between epochs (8,23). This method has the advantage of selectively measuring the amount of time spent in walking, and it also increases the accuracy of energy expenditure predictions across a wide range of activities.

Trail counters.

The number of people using walking trails and sidewalks can be monitored by electronic trail counters. These devices use a variety of technologies, but the basic principle is that they register a count when a person is passing by and record the number of counts in a certain period. Excellent reviews are available of these technologies, and the reader is referred to them for additional detail (11,12).

Active infrared counters transmit a thin pencil beam of radio waves between a transmitter and a receiving unit. They are battery powered and housed in rugged, waterproof cases that are permanently mounted on trees or light poles, typically 1.0 m above ground level. Due to the location of the thin beam, they are unlikely to be triggered by blowing leaves, small animals, and wind-blown vegetation. Passive infrared counters use the same technology contained in motion-sensor lights typically used to turn on outdoor garage lights. Movement of large objects within a wide arc will activate the counter. Geophones connected to seismic sensors within PVC pipes placed underneath walking trails are another means of detecting pedestrians. Another technology is a pressure-sensitive mat, which could be used to record the number of people passing by a particular site.

The main limitation of the above mentioned technologies is that they cannot discriminate walkers from joggers, cyclists, roller skaters, and large animals. Thus, although they provide valid "trail counts," they do not count walkers, specifically. In addition, a given trail user might be counted multiple times. To overcome these limitations, scientists are working on a variety of technologies that use cameras or other devices triggered by active infrared beams to identify and to discriminate among trail users.

Direct observation.

Direct observation of walking behaviors has been used to study time trends in walking at the community level. Although such observations are conducted over a brief period, it is possible to obtain valid counts if the location, season, weather, and time of day are controlled. A long-running survey of this type is the pedestrian counts survey in Christchurch, New Zealand (16). Commissioned by the Christchurch City Council, this survey began in 1957, and it has been conducted continuously at 2-yr intervals ever since. Pedestrian counts were conducted in the central city of Christchurch over two 1-h periods: from 10:30 to 11:30 a.m. and from 2:30 to 3:30 p.m. on a Tuesday and a Thursday in October. Observers are posted at about 60 different locations throughout the city. For sites continuously monitored from 1957 onward, the data show a trend toward a continual decline in pedestrian traffic, which was temporarily reversed in 1982 with the opening of the Cashel/High Street Mall, but after that the decline continued. In the United States, similar trends for a decline in pedestrian traffic were seen in a study commissioned by the Downtown Denver Merchants' Association. The extent to which these declines in walking are due to construction of shops in outlying areas or growth of internet commerce is unknown, but the findings are of serious concern to physical activity researchers, not to mention downtown business owners.

Questionnaires.

Most physical activity questionnaires have items about walking behaviors. For instance, the College Alumnus Questionnaire designed by Paffenbarger et al. (22) asks, "How many city blocks or their equivalent do you regularly walk each day? (Let 12 blocks = 1 mile.)" The US National Health and Nutrition Examination Survey (NHANES) has respondents report how often they walked or bicycled for transportation over the past 30 d and how much time they spent in these activities. In addition, NHANES respondents are asked to recall any specific vigorous or moderate LTPAs (e.g., aerobics, baseball, dancing, walking), how often they performed them, and for how long. However, one of the acknowledged limitations of questionnaires is that they are subjective, and respondents may have difficulty recalling ubiquitous, low- to moderate-intensity walking done throughout the day.

In recent years, it has become common for questionnaires to ask about the total time accumulated in three areas: walking, moderate-intensity physical activity, and vigorous-intensity physical activity. For instance, the International Physical Activity Questionnaire (IPAQ) short form (7) instructs respondents, "Now think about the time you spent walking in the last 7 days. This includes at work and at home, walking to travel from place to place, and any other walking that you might do solely for recreation, sport, exercise, or leisure." IPAQ items include, "During the last 7 days, on how many days did you walk for at least 10 minutes at a time?" and "How much time did you usually spend walking on one of those days?" The intent is to capture all walking performed within the domains of occupation, housework, transportation, and LTPAs (7). However, the IPAQ short form may cause some respondents to overreport walking, moderate activity, and vigorous activity (26).

Transportation surveys are a rich source of information about utilitarian walking and bicycling. For instance, the US National Household Travel Survey (successor to the Nationwide Personal Transportation Survey) uses a 24-h travel diary where people record the details of each trip, including start/end time, destination, distance, and mode of travel (24). Between 1976 and 1995, there was a dramatic decline in the percentage of trips taken by walking in the United States. (Similar declines in active commuting have been noted in Finland and Great Britain.) Thus, it appears that utilitarian walking has declined over the past 30 yr in the United States (24), although this has been partially offset by an increase in leisure-time walking (29).

To summarize up to this point, walking can be measured by many different methods, with varying levels of accuracy. The increasing importance of health-related walking research and of programs promoting healthy walking requires strong research supporting accurate measures of walking. The following sections describe the validation process, which is the process of collecting evidence of the validity of measurement tools.

"WHAT IS VALIDITY?"

Assessment of walking behaviors is important because of the relations demonstrated between walking behaviors and various health and quality-of-life outcomes. As in all research, validity is the most important concept in the measurement of walking. Validity is currently understood as the appropriateness of inferences made from specific measures. The focus of validation is therefore on the intended use and interpretation of test scores rather than on the instrument only. One might infer, for example, that a person does not participate in sufficient walking for maintaining health based on pedometer output. Thus, evidence of the validity of this inference should be judged based on the intended use of the data. A thorough discussion on validity in exercise science research can be found in Mahar and Rowe (18) and in Rowe and Mahar (25).

In human movement sciences, we have traditionally used the conceptualization of validity as presented in psychology measurement (psychometrics). In particular, various versions of the Test Standards, published jointly by the American Psychological Association, the American Educational Research Association, and the National Council on Measurement in Education (1), have influenced the understanding of validity and how validation research should be conducted.

Although the field of psychometrics has contributed much to the development of validity theory, walking research falls under the discipline of kinesmetrics (measurement of human movement). A greater variety of measurement methods are used to assess constructs in kinesmetrics than in psychometrics. Psychologists usually deal with data gathered from questionnaires, whereas exercise scientists not only use questionnaires but also often have well-defined metrics and highly objective measures that do not require a participant to interpret his or her behavior.

An appropriate paradigm for validation in walking research consists of accumulating evidence at three levels or stages, as pictured in Figure 1. First, the definitional stage involves investigation of prior theory and empirical evidence to describe the nature of walking. The confirmatory stage involves investigations that either confirm or disconfirm the definition of walking. The highest level of validation is at the theory-testing stage, where we examine theories of how walking is related to other constructs, including the outcomes and the determinants of walking. The straight arrows pointed upward in Figure 1 emphasize that validity evidence at each stage should be developed hierarchically from evidence gathered at the earlier stages. The curved arrows on the side of the pyramid emphasize the iterative nature of the paradigm, such that information obtained from higher stages may indicate the need to return to previous stages to develop a better understanding of the construct.

FIGURE 1
FIGURE 1:
Three-stage paradigm for validation research in the movement sciences.

Definitional stage.

Before researching or measuring healthy walking behavior, one must first consider the construct definition. The theoretical definition describes the domain of walking (e.g., its characteristics) based on a theoretical understanding of healthy walking behavior. Sources of information might include experts in the field, prior research evidence, or the target population (via interviews, for example). Theoretical definitions include a detailed description of the characteristics of walking behavior that is "healthy" (or perhaps characteristics that differentiate between "healthy" and "nonhealthy" walking). Theoretical definitions may vary depending on the context of the research or purpose of measurement. For example, characteristics of walking that improve health in frail older adults or clinical populations may not describe healthy walking in fit younger adults. Examples of characteristics that may be used to describe walking are speed, intensity, duration, environment, social context, direction, continuity, distance, frequency, rhythm, efficiency, or location.

An operational definition of walking involves describing how walking might be measured. For example, walking can be measured via pedometers, questionnaires, observation, geographical positioning systems (GPS), or more specifically via particular models of equipment (e.g., the Accusplit Eagle pedometer or the Actigraph 7164 accelerometer) or versions of questionnaires (e.g., the College Alumnus Questionnaire or the NHANES survey). Operational definitions also can vary between populations, depending on whether the instrument has been shown to produce accurate scores in a particular population.

When selecting or developing instruments to operationally define walking, one should consider two potential errors (20). Construct under representation means that some aspect of the construct is not measured. For example, a questionnaire used in adults may not include an opportunity to report occupational walking; a pedometer may not record steps taken at slower walking speeds, or a GPS may not record indoor walking. Construct irrelevance means that an instrument records data that are not walking related. Examples are vibrations of a riding mower (recorded as movement by an accelerometer), movement in a vehicle (recorded by a GPS), or socially desirable responses (overreporting the amount of walking on a questionnaire).

An important factor in developing an operational definition of walking is that it should be driven by the theoretical definition, not the opposite way around. One should determine the theoretically defensible definition of healthy walking before deciding how to measure it, not select an instrument or method based on convenience and let the instrument drive theory.

Confirmatory stage.

In the confirmatory stage of construct validation, one conducts studies to test the previously developed definition of walking. Several designs for obtaining confirmatory evidence are presented next.

Factor analysis.

Factor analysis is a statistical method used to confirm the internal structure of a construct. In walking research, factor analysis can be used to confirm that instruments measure various dimensions of walking. For example, Figure 2 illustrates a design in which three instruments (accelerometer, pedometer, and questionnaire) might be used to assess intensity, frequency, and duration. Factor analysis could be used to identify whether the instruments assess the dimensions of walking similarly.

FIGURE 2
FIGURE 2:
Example of a factor analysis design for walking research.

Multitrait, multimethod matrix (MTMM).

MTMM is a method to evaluate convergent and discriminant evidence of validity. Convergent evidence is demonstrated when different measures of the same construct correlate moderately to highly. Discriminant evidence is demonstrated when measures of different constructs do not correlate highly. To conduct an MTMM study, a researcher measures more than one trait with more than one method of measurement. Each measure is then correlated with all other measures. The simplest example involves measuring two traits, each with two methods. A sample 2 (traits) × 2 (methods) MTMM matrix is presented in Figure 3. In this example, the two traits are minutes of low-intensity walking and minutes of continuous moderate-intensity walking in bouts of 8-10 min or longer. The two measurement methods are a pedometer and a questionnaire. The matrix contains correlations that represent reliability coefficients, convergent validity coefficients, and discriminant validity coefficients. In general, results would support validity evidence if reliability coefficients are acceptable, convergent validity coefficients are statistically significant and meaningful, and convergent validity coefficients are higher than discriminant validity coefficients. MTMM allows evaluation of how measurements are affected by extraneous method variance. For example, if two different traits are both measured with the same method (e.g., questionnaires), then the resulting relationship may be, in part, due to the similar measurement methods rather than to a true relationship between the underlying traits. When researchers choose different traits for an MTMM study, they should be guided by the intended use of the instruments. A thorough discussion of the MTMM approach can be found in Mahar and Rowe (18).

FIGURE 3
FIGURE 3:
Example of the multitrait-multimethod matrix.

Regression techniques to obtain criterion-related evidence.

Criterion-related evidence of validity involves examining the relationship between walking and a criterion measure. The two types of criterion-related validity evidence are concurrent evidence and predictive evidence. Concurrent evidence is desired when you want to use a test (called the surrogate measure) as a substitute for the criterion measure. The surrogate measure is often a less expensive or more practical measure than the criterion. The criterion measure in a concurrent validity study should be a highly accurate measure of walking. For example, the accuracy of a pedometer might be examined as a substitute for direct observation of walking.

For predictive evidence, the criterion is measured some time in the future and may not be a criterion measure of the same construct. For example, a researcher may be interested in whether walking behavior is related to future development of hypertension or diabetes.

Terms of relative validity and absolute validity are sometimes used to describe the validity of a walking measure (27). Relative validity is shown by correlations with other criteria. For example, the correlation between healthy steps from a pedometer and minutes of moderate to vigorous physical activity from an accelerometer would be evidence of relative validity. Absolute validity examines whether the amount of activity is accurately estimated. For example, is the number of minutes of walking determined from a questionnaire an accurate estimate of the true number of minutes of walking? Absolute validity is important for researchers interested in determining the proportion of people meeting public health recommendations and in determining the dose-response relationship between walking and health. For many purposes in walking research, one needs to determine whether the measures of walking (e.g., pedometers, questionnaires) yield accurate estimates of the absolute amount of walking performed.

Known difference evidence.

Two different designs can be used to demonstrate known difference evidence of validity. In one design, scores from two populations that are hypothesized to differ in walking behavior (e.g., independent older adults vs older adults in a care facility) are compared. Another approach is to administer an experimental intervention designed to increase walking and to measure walking before and after the intervention. In the first approach, statistically significant and meaningful mean differences between the two groups would provide confirmatory evidence supporting validity. In the latter approach, known difference evidence supporting validity is obtained if the measurement instrument is able to detect a meaningful change in walking behavior. Conversely, if results do not support the hypothesized differences or changes, then confidence in the test scores as measures of the construct would decrease.

Known difference evidence studies are well suited for measures of walking. As with all validation efforts, selection of the different populations or of the experimental intervention should be guided by the intended use of the data. Tudor-Locke (31) presented known difference evidence of pedometers and activity logs for detecting changes in walking behavior consequent to an intervention. Pedometers detected a sizable change in walking due to the intervention, but activity logs did not detect this change.

Theory-testing stage.

After determining via the previous two stages that one is using an appropriate measurement protocol for the intended purpose, one can then use it to test the broader theoretical context of healthy walking. Research conducted at this stage may investigate antecedents of healthy walking or outcomes of healthy walking. Walking is a behavioral experience and also a physiological phenomenon. Theory testing may relate to either or both aspects of walking. Antecedents of walking include the built environment, social support, physical fitness, physical self-efficacy, motivation, fitness, and health. Outcomes of healthy walking include mental well-being, bone health, blood pressure, glucoregulation, mortality risk, and body composition.

Theory-testing research incorporates outcomes and antecedents into designs based on the scientific method to investigate their complex interrelationships. Much of the exercise science research in the past 20 yr has focused on the biological mechanisms that explain why, for example, someone who is regularly active is physiologically different than a sedentary person. Currently, much attention is paid to the behavioral aspects of healthy physical activity to determine what factors explain why someone may be regularly physically active. Although much of this research has included various types of physical activities, the theory underlying walking behavior in particular may be different than other forms of physical activity. For example, the built environment may play a different role in walking behaviors than in other types of physical activities.

An example of the development of theory is mediating-variable research in the psychological domain. This is illustrated in Figure 4. Psychology research in this area investigates whether interventions increase walking directly or via mediating variables. For example, walking programs that involve group walking may increase walking via an increase in mediating variables such as perceptions of safety, feelings of responsibility toward the group, or via an increase in enjoyment of social aspects of walking.

FIGURE 4
FIGURE 4:
A psychological theory-testing paradigm.

In conclusion, validation research into healthy walking behaviors falls under the kinesmetric realm (measurement of human movement). It therefore falls under the three-stage process described here and with additional detail elsewhere (18,25). It is crucial that validation research in the latter two stages (confirmatory research and theory-testing research) is built on directly relevant research at the earlier stages. Theory-testing research involving novel situations, populations, or variables may require returning to the earlier stages to determine that the definitions and instruments are appropriate. Validation research should include reporting of complete, validity-specific information, and readers are directed to a recent source for detailed information (25).

FUTURE DIRECTIONS AND SUMMARY

Given that walking is an important behavior and valid walking assessment is the sine qua non of evidenced-based decisions, what might be recommended for future walking research? As a result of the approximately 50 presentations and over 100 posters presented at the conference on "Walking for Health: Measurement and Research Issues and Challenges," several important questions remain. These are broadly categorized as questions relating to 1) walking and health, 2) measurement of walking behaviors, 3) intervention research, and 4) general research. These categories and representative questions are presented in Table 1. The questions that appear most often are those related to behaviors or interventions. Although there is interest in determining the number of steps necessary for health, the key areas appear to be those related to taking "steps" that result in the adoption of walking behaviors, with sufficient frequency, intensity, and duration to elicit positive health outcomes. Perhaps knowing how much walking is needed for health is the easier question; getting people to make the change is more difficult.

TABLE 1
TABLE 1:
Measurement and research questions raised at the walking for health symposium.

In summary, we have presented measurement instruments, tools, and techniques, information about valid assessment, and important future research topics. Each of these topics, outgrowths from the conference on "Walking for Health: Measurement and Research Issues and Challenges," can influence how one measures walking, what one measures, and the types of research conducted on walking. The success of "Walking for Health: Measurement and Research Issues and Challenges" can influence research agendas and quality of life in countless individuals.

REFERENCES

1. American Psychological Association, American Educational Research Association, and Nation Council on Measurement in Education. Standards for Educational and Psychological Testing. Washington, DC: American Psychological Association, 1999.
2. Balogun JA, Martin DA, Clendenin MA. Calorimetric validation of the Caltrac accelerometer during level walking. Phys Ther. 1989;69:501-9.
3. Bassett DR, Ainsworth BE, Leggett SR, et al. Accuracy of five electronic pedometers for measuring distance walked. Med Sci Sports Exerc 1996;28(8):1071-7.
4. Bassett DR, Ainsworth BE, Swartz AM, et al. Validity of 4 motion sensors in measuring moderate intensity physical activity. Med Sci Sports Exerc. 2000;32(9 suppl):S471-80.
5. Chen KY, Bassett DR. The technology of accelerometry-based activity monitors: current and future. Med Sci Sports Exerc. 2005;37(11 suppl):S490-500.
6. Cooper M. The Inventions of Leonardo da Vinci. New York: The MacMillan Co., 1965, pp. 101-2.
7. Craig CL, Marshall AL, Sjostrom M, et al. International Physical Activity Questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381-95.
8. Crouter SE, Clowers KG, Bassett DR. A novel method of using accelerometer data to predict energy expenditure. J Appl Physiol. 2006;100:1324-31.
9. Crouter SE, Schneider PL, Karabulut M, et al. Validity of 10 electronic pedometers for measuring steps, distance, and energy cost. Med Sci Sports Exerc. 2003;35(8):1455-60.
10. Freedson PS, Melanson E, Sirard J. Calibration of the Computer Science and Applications, Inc. accelerometer. Med Sci Sports Exerc. 1998;30(5):777-81.
11. Gasvoda D. Trail traffic counters: update. Technical Report 9923-2835-MTDC. Missoula, MT: US Department of Agriculture Forest Service, 1999, pp. 1-14.
12. Granner ML, Sharpe PA. Monitoring physical activity: uses and measurement issues with automated trail counters. J Phys Act Health. 2004;1:131-41.
13. Hakim AA, Curb JD, Petrovitch H. Effects of walking on coronary heart disease in elderly men: the Honolulu Heart Program. Circulation. 1999;100:9-13.
14. Haymes EM, Byrnes WC. Walking and running energy expenditure estimated by the Caltrac and indirect calorimetry. Med Sci Sports Exerc. 1993;25(12):1365-9.
15. Hendelman D, Miller K, Baggett C, et al. Validity of accelerometry for the assessment of moderate intensity physical activity in the field. Med Sci Sports Exerc. 2000;32(9 suppl):S442-9.
16. Langford F. Christchurch Central City Pedestrian Counts Survey. Technical Report 01/8. File Number PP/SU/10/3. Christchurch, New Zealand: Christchurch City Council, June, 2001.
17. Lee I-M, Rexrode KM, Cook NR, et al. Physical activity and coronary heart disease in women: is "no pain, no gain" passe? JAMA. 2001;285:1447-54.
18. Mahar MT, Rowe DA. Construct validity in physical activity research. In: Physical Activity Assessments for Health-Related Research. Welk G (Ed.) Champaign, IL: Human Kinetics, 2002, pp. 51-72.
19. Manson JE, Hu FB, Rich-Edwards JW, et al. A prospective study of walking as compared with vigorous exercise in the prevention of coronary heart disease in women. N Engl J Med. 1999;341:650-8.
20. Messick S. Validity. In: Intelligence: Measurement, Theory, and Public Policy (Proceedings of a Symposium in Honor of Lloyd Humphreys). Linn R (Ed.) Urbana, IL: University of Chicago, 1989, pp. 13-103.
21. Montoye HJ, Washburn R, Servais S, et al. Estimation of energy expenditure by a portable accelerometer. Med Sci Sports Exerc. 1983;15(5):403-7.
22. Paffenbarger RS, Wing AL, Hyde RT. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978;108:161-75.
23. Pober DM, Raphael C, Freedson PS. Novel technique for assessing physical activity using accelerometer data (abstract). Med Sci Sports Exerc. 2004;36(5 suppl):S198.
24. Pucher J, Renne JL. Socioeconomics of urban travel: evidence from the 2001 NHTS. Transp Q. 2003;57:49-77.
25. Rowe DA, Mahar MT. Validity. In: Measurement Theory and Practice in Kinesiology. Wood TM, Zhu W (Eds.) Champaign, IL: Human Kinetics. 2006, pp. 9-26.
26. Rzewnicki R, Auweele YV, DeBourdeaudhuij I. Addressing overreporting on the International Physical Activity Questionnaire (IPAQ) telephone survey with a population sample. Public Health Nutr. 2003;6:299-305.
27. Sallis, JF, Saelens BE. Assessment of physical activity by self-report: status, limitations, and future directions. Res Q Exerc Sport. 2000;71:S1-S14.
28. Schneider PL, SE Crouter, Lukajic O, et al. Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk. Med Sci Sports Exerc. 2003;35(10):1779-84.
29. Simpson ME, Serdula M, Galuska DA, et al. Walking trends among U.S. adults: the Behavioral Risk Factor Surveillance system, 1987-2000. Am J Prev Med. 2003;25:95-100.
30. Swartz AM, Strath SJ, Bassett DR, et al. Estimation of energy expenditure using CSA accelerometers at hip and wrist sites. Med Sci Sports Exerc. 2000;32(9 suppl):S450-6.
31. Tudor-Locke, C. A preliminary study to determine instrument responsiveness to change with a walking program: physical activity logs versus pedometers. Res Q Exerc Sport. 2001;72: 282-92.
32. US Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996.
33. Welk GJ, Blair SN, Wood K, et al. A comparative evaluation of three accelerometry-based physical activity monitors. Med Sci Sports Exerc. 2000;32(9 suppl):S489-97.
Keywords:

PHYSICAL ACTIVITY; EXERCISE; KINESMETRICS; AMBULATION; LOCOMOTION

©2008The American College of Sports Medicine