The technology acceptance model

Data Collection in the Palm of Your Hand: A Case Study

Kelly A. Spain and Chad A. Phipps Department of Psychology

Wichita State University

Michael E. Rogers Department of Kinesiology and Sport Studies

Wichita State University

Barbara S. Chaparro Department of Psychology

Wichita State University

Portable computing is an emerging technology that allows computing to occur practi- cally anywhere. Going beyond the typical use as the “pocket-sized organizer,” new methods of using handheld devices are being developed. One new method is to use handheld devices to collect data in the field. The portability of handheld devices allows for data collection in virtually any setting and frees the researcher from the confines of the laboratory. This article describes the process of converting a paper-and-pencil method of data collection to a 3Com Palm Pilot

™ III application. An iterative design

process was used to evaluate the ease of use of the new application. The new applica- tion (a) substantially reduced time to transfer the data to a database, (b) did not inter- fere with the task, (c) allowed the user to accomplish the same tasks as with the paper method while adding functionality beyond the paper method, and (d) was rated as easy to use.

Recent advances in computing technology have led to a dramatic increase in the availability of handheld computing devices such as palm-sized computers, per- sonal digital assistants, personal information managers, and pen tablets. Recent re- search in this area has focused on input techniques such as handwriting recognition (e.g., MacKenzie & Chang, 1999; MacKenzie, Nonnecke, Riddersma, McQueen, & Meltz, 1994) and different (on-screen) soft keyboards (Bohan, Phipps, Chaparro, & Halcomb, 1999; MacKenzie et al., 1994; MacKenzie, Zhang, & Soukoreff, 1999). This

INTERNATIONAL JOURNAL OF HUMAN–COMPUTER INTERACTION, 13(2), 231–243 Copyright © 2001, Lawrence Erlbaum Associates, Inc.

Kelly A. Spain is now at IBM Silicon Valley Laboratory. Requests for reprints should be sent to Kelly A. Spain, IBM, 555 Bailey Avenue, San Jose, CA 95141.

E-mail: [email protected]

research is beginning to provide useful insight for input techniques, and research is arising to provide insight into uses beyond the “personal organizer” and into the realm of data collection (e.g., see Drury, 1987; Johnston, Rushby, & Maclean, 2000; Schwartz, Neale, Marco, Shiffman, & Stone, 1999). Designing for these devices can be difficult due to the small screen size, input techniques (e.g., selection, text entry), navigation, and potential cognitive interference with the main task.

Mobile computing devices provide an opportunity to drastically change the manner in which research data is collected. Computerized data collection has be- come more popular in today’s research settings; however, it is not always possible to use a computer. Desktop computers are not easily moved from one location to another, and although laptop computers provide more mobility, they are costly. Thus, many researchers use paper-and-pencil questionnaires to collect participant responses. However, paper-and-pencil data collection presents several problems to the researcher. First, paper-and-pencil information can be lost or misplaced. Sec- ond, the process of transferring data for analysis may be time consuming and prone to user error. Third, the use of paper can be cumbersome by constantly hav- ing to shuffle papers to ensure the correct sequence of questions is being followed. The use of a mobile handheld computing device to collect data could alleviate these problems. By using a handheld device, the data can be electronically transferred to a computer database for analysis. This process eliminates the need to transcribe from the paper forms; thus the integrity of the data is maintained throughout the process, from collection to analysis. The software could also automatically order and present the questions, allowing the researcher to focus on conducting the as- sessment. The purpose of this study was to replace a paper-and-pencil method of data collection with a method employing a handheld device.

1. BACKGROUND INFORMATION

The Psychology and Kinesiology and Sport Studies (KSS) departments at Wichita State University modified paper-and-pencil fitness assessment forms to work within the confines of the 3Com Palm Pilot™III (Palm). In an attempt to understand the physical capabilities of the aged, the KSS department collects performance data on different physical activities of older (age 60 or older) clients. This information is then compared to a national database, which helps identify clients who are at risk for disabilities. Exercise routines are recommended and the benefits of the exercise can be determined over time by repeating the assessments. Examples of data col- lected include performance on muscle strength, aerobic endurance, flexibility, and motor ability activities.

The KSS department sought to find a way to eliminate the need for manual data entry into a database and to make the data entry as efficient as possible. The assess- ment data are collected in a variety of settings, therefore providing an opportunity to use a handheld device in lieu of a notebook computer or paper-and-pencil forms. The Palm device was chosen due to its small size and low cost. The use of a handheld device allows for direct transfer of the data from the assessments to the database in which they are stored. From this database, these data can be manipu- lated for final analysis.

232 Spain et al.

2. REQUIREMENTS OF THE PROJECT

To successfully shift the data collection process to the Palm application, the follow- ing requirements were established:

• The use of the Palm should not interfere with data collection. The clients and assessments are the focus. We were concerned that the use of the Palm might interfere with the assessment process by either adding time to the process or by hindering interactions between the assessor and clients.

• The use of the Palm should not add any time to the assessment session. The clients are participating for both their own benefit and for research purposes. Their time is valuable and any extra time added to the process may discourage them from participating in or completing the process.

• The Palm forms should be capable of recording data for multiple clients. Cli- ents are often tested in pairs (i.e., couples). With the paper method, the user can flip between papers as each participant’s data is entered on the assess- ment form. The nature of the Palm does not allow for this, but the alternative must be useable.

• Anything that the Palm could do for the user would be considered an added improvement. Without taking control away from the user, we wanted to auto- mate what we could in the process, allowing for more time focused on the as- sessment process itself.

3. METHOD

An iterative design process was employed to determine whether the Palm applica- tion could meet the previously mentioned requirements. A task analysis was con- ducted in which we observed the actual assessment process. The results of the task analysis guided the design process and application functionality.

3.1. Task Analysis

The first step of the task analysis was to view the current paper form. The assess- ment form, including demographic and assessment information, fills approxi- mately 1.5 pages of an 8½" × 11" sheet of paper. The assessor places the form on a clipboard and marks on it using a pencil. Some assessments require timing and unit measurement, so the assessors also carry a stopwatch and ruler with them as needed.

The testing environment. Although the testing environment may differ based on location, the layout is virtually the same. Different assessment stations are arranged to minimize the client’s time spent waiting. Once the older clients arrive at the assessment site, an assessor records their demographic information on a blank

Portable Data Collection 233

assessment form. This information includes name, address, height, and weight data. The form is then given to the client and the client is directed to the next assess- ment station. The testing session lasts approximately 30 to 40 min. This time varies due to the variation in the participant’s individual speed in completing the assess- ments. Before beginning each assessment, the assessor explains the purpose of the test and the requirements of the client. A description of the assessments that are used in the KSS department follows (see also Rikli & Jones, 1999).

Arm Curl Test. The purpose of the Arm Curl Test is to assess upper body strength and endurance. The client is seated on a chair, back straight and feet flat on the floor, with the dominant side of the body close to the edge. A 5- to 8-lb weight is held at the side in the dominant hand. The test begins with the arm in the down posi- tion beside the chair, perpendicular to the floor. The assessor says “Go!,” starts the stopwatch, and the client curls the arm through a full range of motion and then re- turns to the fully extended position in a controlled manner. The assessor counts the total number of correctly completed arm curls, says “Stop!” after 30 sec, then re- cords the score on the assessment form.

Balance Measures Test. The purpose of the Balance Measures Test is to as- sess standing balance. The assessor times the client while the client (a) stands with feet together, (b) stands with feet in the semitandem position, (c) stands with feet in the full tandem position, and (d) stands on one foot. Each position is achieved with (a) the eyes open and head forward, (b) eyes closed and head forward, and (c) eyes open with the neck extended (looking at ceiling). The assessor says “Go!” and starts the stopwatch, and the client begins the test. After 10 sec, the assessor marks whether the client was able to hold the position for that duration. This process is re- peated for all balance measures.

Chair Sit and Reach Test. The purpose of the Chair Sit and Reach Test is to assess lower body (primarily hamstring) flexibility. Starting in a sitting position on a chair, the client moves forward until the client is sitting on the front edge. Keeping one leg bent with foot on the floor, the other leg is extended straight in front of the hip, with heel on the floor and foot flexed. With the extended leg as straight as possi- ble, the client slowly bends forward at the hip joint sliding the hands (one on top of the other with the tips of the middle fingers even) down the extended leg in an at- tempt to touch the toes. The reach must be held for 2 sec. The assessor uses a ruler to measure the number of inches (nearest ½ in.) a person is short of reaching the toes (minus score) or reaches beyond the toes (plus score). The assessor takes measure- ments on each leg, records the data in the margins of the form to determine which leg is the “superior” leg (the leg with the best score), then records two trials on the superior leg.

234 Spain et al.

Chair Stand Test. The purpose of the Chair Stand Test is to assess lower body strength and endurance. The test begins with the client seated in the middle of the chair, back straight, and feet approximately shoulder width apart and flat on the floor. Arms are crossed held against the chest. The assessor says “Go!,” starts the stopwatch, and begins counting as the client rises to a full stand (body erect and straight) and returns back to the initial seated position. The client is encouraged to complete as many full stands as possible within a 30-sec time limit. After 30 sec, the assessor says “Stop!,” totals the number of stands executed correctly, and records that number to the assessment form.

Eight-Foot Walk Test. The purpose of this test is to assess walking speed. The client is timed while walking on an 8-ft path. Infrared sensors interfaced with a stop- watch detect when the client starts and finishes the task. The client performs the task at normal walking speed. The client also performs the task using a tandem walk (placing the heel to the toe on each step). When the client finishes the walk, the asses- sor records the time from the stopwatch and resets the stopwatch. This process may be repeated two or three times.

Scratch Test. The purpose of the Scratch Test is to assess upper body (shoul- der) flexibility. In a standing position, the client places the preferred hand over the same shoulder and reaches as far as possible down the middle of the back, palm down and fingers extended. The hand of the other arm is placed behind the back, palm up, reaching up as far as possible in an attempt to touch (or overlap) the ex- tended middle fingers of both hands. The assessor measures the distance of overlap, or distance between the tips of the middle fingers to the nearest ½ in. Minus scores (–) are given to represent the distance short of touching middle fingers; plus scores (+) represent the degree of overlap of middle fingers. The assessor measures the cli- ent on both sides and records those preliminary scores in the margin. The assessor then takes two measurements on the superior limb and records those scores on the assessment form.

Six-Minute Walk and Two-Minute Step-in-Place Test. The purpose of the Six-Minute Walk Test is to assess aerobic endurance. The test involves assessing the maximum distance the client can walk in 6 min along a 50-yard course, marked into 5-yard segments. Clients continuously walk around a measured lap throughout the 6-min period, trying to cover as much distance as possible. The assessor says “Go!,” starts the stopwatch, and begins counting the number of completed laps as the cli- ent walks (not runs) as fast as possible the marked distance along the course as many times as they can within the time limit. If necessary, clients may stop and rest, sitting on chairs provided, then resume walking. The assessor counts the number of laps completed for each client. After 6 min have elapsed, the assessor says “Stop!” and records the total number of completed laps on the assessment form.

Portable Data Collection 235

The purpose of the Two-Minute Step-in-Place Test is to assess cardiovascular and physical endurance in older adults who are unable to complete the 6-min walk. The proper (minimum) knee-stepping height for each client is at a level even with the midway point between the patella and the iliac crest. To monitor correct knee height when stepping, a ruler is attached to a chair with masking tape marking the proper knee height. The assessor says “Go!,” starts the stopwatch and begins counting completed steps as the client begins stepping (not running) in place, com- pleting as many steps as possible within the time period. As soon as proper knee height cannot be maintained, the assessor asks the client to stop—or to stop and rest—until proper form can be regained. Stepping may be resumed if the 2-min time period has not elapsed. If necessary, one hand can be placed on the table or chair to assist in maintaining balance. After 2 min have elapsed, the assessor says “Stop!” and records the total number of steps completed.

During either of these assessments, the body mass index (BMI) is calculated. BMI (Revicki & Israel, 1986; Smalley, Knerr, Kendrick, Colliver, & Owen, 1990) is an index of height and weight that is indicative of various disease risks including car- diovascular disease and Type II diabetes. Clients are provided their score, indicat- ing whether they are high or low risk. A score of 25 or higher is considered high risk. The calculation for BMI is

The assessor uses a calculator to determine each client’s BMI.

Up and Go Test. The purpose of the Up and Go Test is to assess physical mo- bility—involving speed, agility, and dynamic balance. The test begins with the cli- ent fully seated in the chair, hands on thighs, and feet flat on the floor. The client is told that this is a “timed” test and that the object is to walk as quickly as possible (without running) around a cone and back to the chair. The assessor says “Go!” and the client gets up from the chair, walks as quickly as possible around the cone, and returns to the chair. When the client has returned to the chair, the assessor stops the stopwatch and records the time elapsed. This assessment is conducted twice.

When all assessments are complete, the forms are collected to be later entered in a computer database. Two people are involved in this part of the process; one per- son reads the data from the assessment form while the other person enters the data in the database. This process was used to help reduce data entry errors. One client’s data is entered into the database in approximately 5 min.

3.2. Interface Design and Functionality

The task analysis revealed several important issues. As noted before, for this project to be successful it was imperative that the Palm application be as flexible and func-

236 Spain et al.

2

Weight in kilograms (Height in meters)

tional as the paper method. Based on the existing paper forms, field observations, and the task analysis, a prototype was developed using Satellite Forms 3.0 (Pumatech, San Jose, CA). Each assessment is displayed on one form. Users navi- gate the forms via back and next buttons, in a wizard-like fashion. This design was chosen based on the assumption that users would go through the assessments (and therefore, the assessment forms) in a specified order.

Scratchpad. It was observed that during the Chair Sit and Reach Test and Scratch Test the assessor would take measurements for both right and left limbs, then write those values in the margin of the paper form. This data would not be recorded to the database, but was helpful to the assessor for determining, and re- membering, which was the superior limb. Subsequent measurements were taken on the superior limb only. This led to designing a scratchpad area that allows the assessor to record the initial measurements (see Figure 1). As with the paper method, these data are not transferred to the database, but aid the assessor dur- ing the assessments.

Timer. Several assessments require a number of actions to be counted during a given time period (e.g., the Chair Stand Test consists of the number of stands in 30 sec), whereas others require actions to be timed (e.g., the Up and Go Test consists of the time to stand from a chair, walk around a cone, and return to the chair). Built-in timers were designed in those forms with assessments requiring timing (see Figure 2 for an example). This eliminates the need for a separate timer or stopwatch. For the timed assessments, the time data are input directly from the timer field into the data fields. For the assessments requiring set time periods, the Palm application emits a “beep” at the end of the duration.

Portable Data Collection 237

FIGURE 1 Scratch test assessment form with scratchpad area.

BMI calculator. To address the issue of calculating the BMI by hand, a BMI cal- culator function was designed in the demographic form where height and weight data are entered. When the client’s height and weight data are available, the BMI is calculated by tapping a graphical button on the screen.

4. USABILITY TESTING

Three usability tests were conducted to evaluate the new application interface. All three tests were conducted in the assessment environment with different users.

4.1. Users

For each of the three usability tests, the users were recruited from the KSS depart- ment. All users had experience with the assessment process and had administered them using the paper method. For the different usability tests, the users had differ- ing degrees of experience with the Palm application, described in detail for each us- ability test. A total of eight users participated in the usability tests. We wanted the users to be familiar with the assessment process before the usability testing and therefore used all available and qualified graduate students in the KSS department.

4.2. Procedures

Before the test, each user was given a 15-min instruction period on interacting with the Palm device. This consisted of demonstrating the soft keyboards for alphanu- meric data entry and the graffiti tool for numerical data entry. The graffiti tool (Mac-

238 Spain et al.

FIGURE 2 Up and Go assessment form with timer.

Kenzie & Zhang, 1997) is an alphabetic language written one character at a time, us- ing the stylus. The alphanumeric characters resemble that of the regular printed alphabet. The majority of the data collected are numerical data. We felt the graffiti tool lent itself well to this type of entry, as the numbers for the numerical graffiti tool closely resemble printed numbers.

Users were told of the purpose of the test. Because the test was conducted in the field, the use of video cameras was impractical. Therefore, an observer followed the user, recording any comments or problems, and prompting the user for comments if necessary.

Tasks. The task in all usability tests was for the user to perform the assess- ments in the normal manner, except to enter data in the Palm application. On com- pletion of all assessments, we asked the users to rate the Palm application on how easy or difficult it was to use and to state their willingness to switch from the paper method to the Palm application.

4.3. Test 1: Functionality

The purpose of this test was to determine if the users were able to interact with the Palm application with 15 min of training and to test the additional functionality of the BMI calculator, timer, and scratchpad.

Users. Three users completed Test 1. All three users were familiar with the as- sessments but unfamiliar with the Palm device. Users familiarized themselves with the Palm as described earlier.

Results. All three users were able to complete the task. One user exhibited dif- ficulty entering data in the Palm. Data is entered either with the soft keyboard or graffiti tool, but this user attempted to enter numerical data directly in the data fields. For the first three forms the user tried this without success. Once the user was alerted that data could only be entered via the soft keyboard or graffiti tool, the user used the graffiti tool. All three users were able to use the scratchpad fields without errors. No user tried to enter the preliminary data in the data fields.

All users made general positive comments about the use of the Palm for data. One user commented, “This is a great idea to step into technology with the assess- ments.” Specific comments include one user stating, “The calculation for the BMI saves time and I don’t have to do it myself.” Another user stated, “This [BMI calcu- lator] button is great. Just one touch and the number appears!”

One user specifically had problems navigating the screens stating, “I just saw the Arm Curl page, how do I get back?” All three users showed some difficulty navigating the different forms as evidenced by clicking the back and next buttons repeatedly. The assessment forms did not appear in the same order the assessments

Portable Data Collection 239

were being conducted. This rendered the back and next buttons insufficient for navigation. One user continued clicking forward to the end then back to the begin- ning before finding the desired form. When asked to rate the ease of use of the Palm application on a scale of 1 to 5, with 1 being difficult to use and 5 being easy to use, two ranked it as a 3, and one ranked it as a 4. All three users stated they would be will- ing to switch to the Palm application.

Conclusions. The major finding of Test 1 for improving the design related to navigation. As stated earlier a wizard-like approach was taken because the assess- ments are conducted in a particular order. However, during this test we realized this might not always be the case and designing it thus made the program inflexible. A drop-down menu listing all but the current form was designed in all forms so users could jump to other forms. This was tested in Test 2.

4.4. Test 2: Navigation and Multiple Client Measurements

The goal of Usability Test 2 was to evaluate the addition of the drop-down naviga- tion menu and to determine the effectiveness of recording more than one client’s data at the same time. The Palm provides a hard-wired scrolling mechanism to scroll through different records as well as a menu system to select the next or previ- ous record. Therefore, shifting between different clients was not a feature we de- signed in the application, as there were other alternatives. We demonstrated how to use both the scroll buttons and the menu to switch between client forms.

Users. Three users completed Test 2. All users were familiar with the assess- ment process. One user was familiar with the Palm device. The other two users were given time to familiarize themselves with the Palm device as described earlier.

Results. Users navigated using both back and next buttons and the drop-down menu without stating or exhibiting any difficulties. Users had difficulty with the main task of this test: to record two clients’ data at the same time. Two users entered client’s data on the wrong form, and when asked about it, they replied they were not aware of the error. When asked to rate the ease of use of the Palm application on a scale of 1 to 5, with 1 being difficult to use and 5 being easy to use, one user gave the pro- gram a 3, whereas two users gave it a 4. All three users stated they were willing to switch to the Palm application.

Conclusions. No problems arose regarding navigation. The Palm application was rated as easy to use by two of the participants, and again all three users stated they would be willing to use it in place of the paper forms. The difficulty in differen-

240 Spain et al.

tiating between clients’ forms led to adding a name field to all forms. This design change was tested in Test 3.

4.5. Test 3: Validation

The goal of this test was to evaluate recording two client’s data simultaneously and to identify any other issues the users may have.

Users. Two users participated in this test. Both users were familiar with the as- sessment process, but unfamiliar with the Palm device. Users familiarized them- selves with the Palm as described earlier.

Results. Users were successful at recording data for two clients without trying to enter data for the wrong person. No other problems were noted. Comments about the Palm application were all very positive. One user stated that she thought this was a “great alternative.” Other comments included the added functionality provided by the BMI calculator and timer. When asked to rate the ease of use of the Palm application on a scale of 1 to 5, with 1 being difficult to use and 5 being easy to use, both users rated the application as a 5. When asked, users stated they would feel comfortable using the Palm application in place of the paper method.

Conclusions. Placing the name field on each form made it easier for the users to know on which client’s form the data was being recorded. The Palm application was rated as easy to use, and the two users stated they were willing to switch to this method. The lack of problems the users encountered in this usability test reinforced the design of the Palm application.

5. DISCUSSION

The Palm application was developed to replace the paper method, not to determine if one was superior in data entry. Therefore, the two methods were not systemati- cally compared. A typical assessment takes approximately 30 to 40 min with the pa- per method. Assessments with the Palm application took this same length of time. This was not surprising considering that the length of the assessment process often depends on the client’s physical health: The types of measures taken rely on the speed at which the client completes the assessments. We saw the greatest timesav- ing during data transfer. Whereas it takes 5 min to manually transfer one record to a database, it takes a few seconds to electronically transfer one record to a database via the HotSync function. This relates to the Palm application being 100 times faster than the paper method for data transfer.

Portable Data Collection 241

Usability tests of the Palm application were conducted throughout the design process. During these tests, the users were generally unfamiliar with the device but were familiar with the fitness assessment process. The unfamiliarity with the de- vice did not seem to cause major problems with the users. Although the paper method was familiar to all the users, they stated that they were comfortable and willing to switch to the Palm application exclusively. This finding emphasizes the success of the application. With the users feeling comfortable with the Palm appli- cation it is unlikely that the device would become an issue and interfere with the data collection process.

The Palm application met our goals: It did not interfere with the assessment pro- cess; it did not add time to the assessment process; users were able to record data for multiple clients; functions were added to automate and aid in several tasks; and users rated it as easy to use. We found the Palm device to provide an excellent plat- form for field data collection by providing a portable and relatively cheap data col- lection tool. This application is now being used in the KSS assessment process and has replaced the paper-and-pencil forms.

Several issues arose while completing this project. The first and most drastic was the limitations of the device itself. Specifically, it was challenging to design for such a small screen space (160 × 160 pixels). Navigation was also important in this de- sign. Initially, the only type of navigation was a wizard-like approach involving back and next buttons. The first usability test revealed that the order of the assess- ments is dependent on the assessment location. The Palm application needed to be more flexible. The addition of the drop-down menu allowing the user to jump to other assessment forms alleviated navigation problems.

Our users were generally unfamiliar with the device but were open to using the Palm application even when they encountered difficulties with it. This may be due to their knowledge of the goal of the project (replacing the paper forms). It is possi- ble that less willing users may have reacted to the new technology differently.

REFERENCES

Bohan, M., Phipps, C. A., Chaparro, A., & Halcomb, C. G. (1999). A psychophysical compari- son of two stylus-driven soft keyboards. Proceedings of Graphics Interface ’99 (pp. 92–97). Kingston, Ontario: Canadian Human Computer Communication Society.

Drury, C. G. (1987). Hand-held computers for ergonomics data collection. Ergonomics, 18, 90–94.

Johnston, N. A., Rushby, N., & Maclean, I. (2000). An assistant for crew performance assess- ment. The International Journal of Aviation Psychology, 10, 99–108.

MacKenzie, I. S., & Chang, L. (1999). A performance comparison of two handwriting recognizers. Interacting with Computers, 11, 283–297.

MacKenzie, I. S., Nonnecke, B., Riddersma, S., McQueen, C., & Meltz, M. (1994). Alphanu- meric entry on pen-based computers. International Journal of Human–Computer Studies, 41, 775–792.

MacKenzie, I. S., & Zhang, S. (1997). The immediate usability of Graffiti. Proceedings of Graph- ics Interface ’97 (pp. 129–137). Toronto, Canada: Canadian Information Processing Society.

242 Spain et al.

MacKenzie, I. S., Zhang, S. X., & Soukoreff, R. W. (1999). Text entry using soft keyboards. Be- havior and Information Technology, 18, 235–244.

Revicki, D. A., & Israel, R. G. (1986). Relationship between body mass indices and measures of body adiposity. American Journal of Public Health, 76, 992–994.

Rikli, R. E., & Jones, C. J. (1999). Development and validation of a functional fitness test for community-residing older adults. Journal of Aging and Physical Activity, 7, 129–161.

Schwartz, J. E., Neale, J., Marco, C., Shiffman, S. S., & Stone, A. A. (1999). Does trait coping ex- ist? A momentary assessment approach to the evaluation of traits. Journal of Personality and Social Psychology, 77, 360–369.

Smalley, K. J., Knerr, A. N., Kendrick, Z. V., Colliver, J. A., & Owen, O. E. (1990). Reassessment of body mass indices. American Journal of Clinical Nutrition, 52, 405–408.

Portable Data Collection 243

Copyright of International Journal of Human-Computer Interaction is the property of Taylor & Francis Ltd and

its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's

express written permission. However, users may print, download, or email articles for individual use.