Conducting an Interactive Survey of Household Weekly Activities via
Contents
1. DOT Kurani K S and R Kitamura 1996 Recent developments in the prospects for modeling household activity schedules A report prepared for the Los Alamos National Laboratory Institute of transportation studies University of California Davis California 1996 Lawton T K 1996 Activity and time use data for activity based forecasting Paper presented at the Activity Based Travel Forecasting Conference New Orleans June 2 5 1996 Lee M Doherty S T Sabetiashraf R and McNally M G 2000 CHASE An internet Household Activity Scheduling Elicitor survey Presented at the 79th Annual Transportation Research Board Meeting Washington D C January 9 13 2000 Recker W W McNally M G and Root G 1986 A model of complex travel behavior Part I Theoretical development Transportation Research A 20 4 307 318 Recreation Eat Sleep Shower Night sleep Entertainment Jogging biking roller skating Shower dress pack Fitness center Dinning out at restaurants Golf Snack Spectator sports Breakfast Bars Lunch Movies in theaters Dinner at home Watching videos Work School Work Regular TV programs School only if you are a student Browsing Web sites Meal preparation Relaxation Rest Cleaning Maintenance at Hobbies at home crafts home gardening and others Pick up drop off kids Pleasure driving Pick up drop off others Visiting Attending to children at home Hosting visitors Phone email over 10 Filling o2. If yes did you find answers to your questions 3 When you were asked to enter your frequently visited locations were you able to select most of the locations you often visit from the list provided 4 When you were asked to select your typical activities were you able to select most of the activities you often do from the list provided Part II Weekly Calendar 5 On the first Sunday a Were you aware that you should not intentionally plan and enter more activities than you had already planned prior to beginning the program b Do you think you follow this principle reasonably well 6 Were you aware that if there were any details about an activity that you did not know you should leave the field as a I don t know response 7 Were you aware that if there was an activity you know you would do in the survey week but the exact date had yet to be determined you should enter the activity in the Any Day column Part III Daily Calendar 8 Were you aware that if you ever walked rode a bicycle took a bus or drove a car for over 5 minutes to get to your next activity you should enter this travel time between activities to your Final Schedule 9 When there were boxes popping up to ask you for reasons were you aware that you can select No specific reasons or I was not thinking about anything if indeed that was the case 10 Did you find the Daily Calendar easy to use when
3. Activity Diary begins on a Sunday and ends on the very next Sunday The Initial Interview is proceeded the first time you run the program It should be finished no later than the Sunday evening when the recording of Weekly Activity Diary begins It is advised that you install and begin to run the Initial Interview before this Sunday so the amount of time you spend on entering data can be reduced The recording of the Weekly Activity Diary begins on a Sunday evening you have to complete the Initial Interview Each night you should run the program anytime from 6PM to 12AM midnight If you prefer to run the program after midnight you can do so from Monday to Saturday But on both the beginning and ending Sundays you must run the program before midnight On the first Sunday evening you will be asked to enter the activities for the upcoming week from Monday to the next Sunday On Monday evening you will be asked to update what you have actually done throughout the day Also you will be asked to update changes you have made to your activity plans for the rest of the week i e Tuesday to Sunday You will need to repeat this process for every evening of the surveying week Upon completion of the survey you will receive your financial compensation of 25 for a single person or 50 for married couples Should you decide to terminate your participation in the study before completion you will receive a partial compensation This amount will be de
4. do this activity at this location An Any Day activity was executed or it has been scheduled on a certain day in the week Table 2 Decision Tracing Dialogs Task Initial Sun Mon Tue Wed Thu Fri Sat Sun Total Interview Ave 18 00 22 37 20 23 20 40 13 63 13 23 13 40 11 73 15 07 146 87 Time Min Table 3 Breakdown of Data Entry Time by Day INumber of Entries per Description Entries Person Percentage ork School 119 9 ___ 2 08 Household Obligation o6 ju 28 57 Recreation Entertainment 6 PI R0 Eat sleep shower dinning out at restaurants Table 4 Activities Entered on the Beginning Sunday Activity Categories Activities Number of Entries Shopping Major and Minor Grocery 8 House wares clothing personal items Recreation and Fitness center hobbies at 5 Entertainment home crafting gardening and others pleasure driving Services and Errands Banks post offices gas 5 stations hair cut stores Household Obligation Cleaning and maintenance 5 Social Family dinner friends 3 gathering phone email over 10 minutes Table 5 Any Day Activities Questions Number of Yes Number of No Missing Part I Initial Interview 1 Did the survey questions provide clear instructions on what and how to enter your responses 9 0 2 Did you ever click on the Help button
5. is 26 the youngest is 20 the eldest is 29 All the graduate students are part time employees of UCI Two of the spouses are full time home makers each has a child under 1 year old The lowest education level of the spouses is associate degree of a 2 year college Breakdown of data entry time The average data entry time is presented in Table 3 It has to be noted that some households skipped a day in the week Recording of the skipped activities was later completed on the very next day In table 3 entry time of these make up sessions was tabulated under the day it was skipped Overall the entry time decreased as respondents became familiar with the program Among couples the second person to enter data usually spent less time than the first person This may be an indication that couples learned from each other how to run the program Activities Entered in the First Sunday Evening A total of 371 activities or an average 25 activities per person were entered to the Weekly Calendars on the beginning Sunday evening see Table 4 These were activities that participants knew they would do throughout the surveying week In the pilot respondents were asked not to enter basic activities such as night sleep shower and regular meals to their Weekly Calendars These activities are usually performed everyday at a regular time It was decided that the typical time windows of these activities be inquired in the Initial Interview to reduce the number of e
6. of the interaction among household members on each member s schedule In the past decade models of household activity scheduling became a focal point among researchers as the objectives for travel demand modeling changed Currently emphasis of transportation policies is travel demand management i e efficiently using existing facilities to fulfill people s needs for activities rather than increasing the means for travel New policies require the evaluation of how people would temporally and spatially adjust their travel behavior when the supplies in the activity transportation systems were changed Karash and Schweiger 1994 In light of these policies the modeling of household activity scheduling as interdependent decisions becomes more important to practical demand modeling than ever before There is a large body of literature e g Recker et al 1986 Axhausen and Garling 1991 G rling et al 1994 and Kurani and Kitamura 1996 in the field of activity based analysis noting that the inefficiency of existing models is resulted from the lack of more in depth research on the nature of human activity behavior Over decades the community of activity analysis had to rely on observed activity travel diaries for research Researchers were not able to explore the dynamics of activity scheduling since the travel diaries only recorded the outcome of decision making not the process Axhausen 1995 noted that more data need to be collected from h
7. other answers to these questions were provided by the person who acted as the correspondent to the survey administrator Currently opinions of the second person were not sought The responses in Table 6 clearly demonstrates that interviewees were able to follow the program s guidance to complete designated data entry tasks SUMMARY AND FUTURE RESEARCH A new computer program REACT has been developed to collect data on household activity scheduling process The program is currently being tested in a pilot study in Irvine California This paper describes the program design and results from a subset of pilot samples The design of the program greatly reduces the cost of administrating an interactive household survey by allowing respondents to use their own computers to input data and upload them to the survey administrative server Significant advancements were also made to trace decisions involved in the scheduling process Further analysis of the resultant data is pending until a target sample size of 50 households is reached However initial analysis on a subset of the sample validated the program s capability of guiding participants to complete data entry tasks on their own Based on this preliminary finding REACT represents an ideal platform to realize Lawton s 1996 vision of a computerized household survey The Decision Tracing Dialogs can be easily adapted to stated response or stated preference questions In addition the sample size of su
8. the survey administrative server via FTP The ID and date on each package will enable investigator to monitor the progress of each participating households and examine the validity of the data The entire Send Data process is performed automatically without user intervention When it finishes users will receive notification to exit the program for today PILOT STUDY Beta Testing and Survey Procedure The prototype of REACT has gone through a month from late February to the end of March 2000 of beta testing with 9 households The testing was intended to detect errors and inefficiencies of the program Data obtained will not be analyzed in any way other than detecting errors During this period 2 to 3 households were recruited each week to participate in the testing following the same procedure intended for the official pilot study First flyers see Appendix were sent to households in the study area two student communities around University of California Irvine In addition a Web site www its uci edu react was established to give potential participants a brief introduction of what questions and tasks they should be expecting from the survey Participant were financially compensated A single person received 25 and a couple received 50 Interested participants would contact the survey administrator by phone or email A package containing REACT program CD and other administrative materials i e survey consent forms instruction on when t
9. Figure 1 In addition before the calendar shows up there are always highlighted messages in the automatic slide show warning respondents not to intentionally plan and enter more activities than those they have known before they run the program Each adult member in the household will have a separate calendar to work on Interviewees enter known activities to the calendar by activating the Activity Information entry form see Figure 11 Note that users can leave details of an activity e g time location and duration as I don t know if they are not sure about them at the moment Activities intended for the future appear in a stack of boxes showing known details The Any Day list of activities in Figure 8 is for user to enter activities that will be done in the week but the date has yet to be determined The Any Day column will remain on screen whereas the remaining days can be viewed by scrolling when needed If a respondent enter a specific activity that the survey administrator is interested in investigating why the activity will be done on a particular day a dialog box will show up to request this information from interviewees see question 5 in Table 2 Daily Calendar At the end of each day in the week interviewees will be taken to the Daily Calendar as soon as they turn on the program On this calendar see Figure 9 activities indicated for the current day are listed in a stack of boxes on the Tentative Activities column on the
10. UCI ITS AS WP 01 1 Conducting an Interactive Survey of Household Weekly Activities via Internet Preliminary Results from a Pilot Study UCI ITS AS WP 01 1 Ming S Lee Ramesh Sabetiashraf f Sean T Doherty j Craig R Rindt Michael G McNally 1 Institute of Transportation Studies University of California Irvine Irvine CA 92697 3600 U S A Reazon Systems 3663 S Bear Street Santa Ana CA 92704 U S A Department of Geography amp Environmental Studies Wilfrid Laurier University Waterloo Ontario N2L 3C5 Canada January 2001 Institute of Transportation Studies University of California Irvine Irvine CA 92697 3600 U S A http www its uci edu Lee et al Conducting an Interactive Survey of Household Weekly Activities via Internet Preliminary Results from a Pilot Study Ming S Lee Ramesh Sabetiashraf Sean T Doherty Craig Rindt and Michael G McNally Institute of Transportation Studies University of California Irvine Irvine CA 92697 3600 USA Tel 949 824 6571 Fax 949 824 8385 Email mingshel uci edu mmcnally uci edu crindt translab its uci edu Reazon Systems 3663 S Bear St Unit K Santa Ana CA 92704 USA TEL 714 556 1040 Email sabeti reazon com Department of Geography amp Environmental Studies Wildrid Laurier University Waterloo Ontario Canada N2L 3C5 Email sdoherty wlu ca Abstract The primary goal of activity based models is a fundamental examinati
11. add one by themselves Activities such as night sleep shower and other essential daily activities are automatically selected as defaults to shorten data entry time The selected activities will be written to REACT s central database Within the database survey administrators can designate what attributes of a specific activity they are interested in investigating If users select these activities they will have to enter on the following subordinate forms the attributes that the investigator designated b Activity Frequency On this sub form users will enter typical frequencies for the activities designated by the investigator For example major grocery over 10 items 2 times every week c Activity Duration This form is designed to collect information about the typical duration of certain activities For example jogging 30 minutes d Activity Time and Day Figure 5 This form asks users the typical time windows on which they would do certain activities Users have to indicate if a specific activity usually occurs on a fixed time and day If it does occur on fixed time and days they have to enter the fixed days and time windows If not they will select the days and time on which it could occur e Activity Locations Figure 6 On this form users will select a list of locations at which they can perform a certain activity f Involved Persons Figure 7 If other persons occasionally do this activity for the current interviewee so h
12. alifornia Alberts Heritage Plaza on Walnut and Culver Ralphs B Heritage Plaza o Ralphs Burger King A Irvine Village Center on Irvine Center D and Jeffrey Irvine Dance Academy Ple 17 Back HE Figure 6 Typical Activity Locations Entry Form Initial Interview Click at the yellow page to get online help EZE ae VISER C People outside my household People outside my household Next gt meh Figure 7 Involved Persons Entry Form Weekly Calendar for John File Edit Help Tue 1 11 00 Start End Dur 0 09 00 AM 05 00 PM Dur 48 Start 2 End Dur 09 00 AM 05 00 PM Dur 480 Figure 8 Weekly Calendar Daily Calendar Edit Help Final Schedule Tentative Schedule Mon 1210700 Any Day Activities ViajOf a Cer View Calendar Start End Dur 0 Start End Dur 0 09 00 AM 05 00 PM Dur Figure 9 Daily Calendar w Read me Beginning Sunday On the beginning Sunday evening please tell us what activities you are going to do for the coming week i e from Monday morning till next Sunday evening On the background is the Weekly Calendar on which you are going to enter your activities Please note at this moment we do NOT want you to enter your hour by hour schedule for the coming week Just tell us the activities you KNOW FOR SURE that you will be doing on each day of the week These activities ma
13. approach is suitable for screen intensive applications like CHASE that demands fast interactive data entry Data transferring can be performed at the login and logout so data entry time is not affected by the speed of Internet connection In addition stability and performance of the application will not be damaged by the occasional network traffic congestion The downside is that clients will have to install the program in their computers However this can be amended by streamlining the self installation process and packaging un installation utility with the program Lee et al 5 Program Features REACT is implemented as a fat client system because it provides the desired speed and stability The program is written as a windows application with Internet connectivity and packaged with self executing installation and un installation files It inherited the survey structure of CHASE by dividing the process into two self completing data entry stages Initial Interview and Weekly Activity Diaries Initial Interview is a series of questionnaires intended to collect the same information as those collected in CHASE s up front interview The process of activity scheduling is collected in the Weekly Activity Diaries Participants of the survey have to install the program and finish the Initial Interview before the Sunday evening when the recording of Weekly Activity Diaries begins Data collected from the interview will be automatically written to the main da
14. ch a survey can be augmented by asking participants who have compatible PCs to install the program on their own Field personnel with laptop computers can then collect data from individuals who do not have PC or do not know how to use computers Such a combined sampling approach should result in a moderate sample size with a controlled bias on education or income level symbolized by computer literacy Lee et al 12 REFERENCES Axhausen K W 1995 The data needs of activity scheduling models Paper presented at the International Conference Activity based Approaches Activity scheduling and the analysis of travel patterns Eindhoven University of Technology The Netherlands May 25 28 1995 Axhausen K and G rling T 1992 Activity based approaches to travel analysis Conceptual frameworks models and research problems Transport Reviews 12 323 341 Doherty S T and E J Miller E J 1997 A computerized household activity scheduling survey Paper presented at the IATBR 1997 conference Austin TX September 1997 Garling T Kwan Mei po and Golledge R G 1994 Computational process modelling of household activity scheduling Transportation Research B 28B 5 355 364 Karash K H and C Schweiger 1994 Identification of transportation planning requirements in federal legislation Final report prepared for John A Volpe National Transportation Systems Distributed in cooperation with Technology Sharing Program U S
15. e she can do something else the interviewee needs to select them from a provided list i e it contains the interviewee s household members and an extra option labeled as People outside my household In addition if the interviewee usually does this activity with other persons they also need to be indicated from that list 7 Transportation Modes Users will select the modes they use from a list of local transportation modes including walk ride with others van pool car local bus bicycle and commuter rail Lee et al 7 Weekly Activity Diaries Weekly Activity Diaries containing two separate graphical user interfaces for interviewees to record their activity scheduling process Weekly Calendar Figure 8 is used to record activity intentions before they occur and Daily Calendar Figure 9 is to keep track of what activities actually occur during a day Instructions on how to enter data to the calendars are given to users through a automatic slide show Figure 10 that launches itself when the program is turned on after Initial Interview is completed Weekly Calendar On the Sunday evening when the surveying week begins users will be taken to an empty Weekly Calendar on which they will enter activities they know they will do on each day of the coming week To minimize the potential fill up bias i e encouraging unnecessary planning by showing respondents time tables a Weekly Calendar does not contain a time scale as CHASE see
16. l Schedules to the current time they will be taken back to the Weekly Calendar and asked to review the activities for the subsequent days and update them if changes have been made e g a new appointment has been scheduled for one of these day Before users see their Weekly Calendars the automatic slide show will launch again to warn interviewees not to intentionally plan and change activities if they have not learned or thought about the changes before they run the program When changes are made the question boxes will show up again to trace the reasoning process The process of updating Daily and Weekly Calendars will continue everyday until the coming Sunday when the survey ends Decision Tracing Dialogs REACT is implemented with mechanisms intended to trace decisions involved in everyday activity scheduling When a respondent manipulates a specific activity record in a certain way a series of dialog boxes will show up to trace the decision process underlying the manipulation Nine different questions will show up as soon as interviewees change or add a specific activity record in their Daily Calendars Table 2 lists each specific data manipulation and its corresponding question In the current implementation no question will be issued to any addition or changes involving basic activities such as night sleep shower and regular meals except for dinning out at restaurants If a respondent performed more than one manipulation on a single acti
17. left hand side Any Day activities are listed in the same fashion on the right hand side The Final Schedule for the current day is placed in the middle with a specific time scale It is noted that displaying the current day s schedule along a standard time line as in the original CHASE survey is not subject to the fill up bias at this stage as these activities have already been executed Users first review the Tentative and Any Day activities and identify those that were actually executed in the current day Then they need to activate the Activity Information form to specify all the details about these activities Immediately after they enter all the details a series of dialog boxes containing questions may show up if an activity was executed in a way different from their intention see the section of Lee et al 8 Decision Tracing Dialogs for details When questions in the dialog boxes are answered they can move this activity to the Final Schedule by pressing down the left from Tentative to Final or right from Any Day to Final arrow buttons Interviewees will continue to finish their Final Schedule by adding activities and travel events see Figure 12 that actually occurred but were not indicated as either the Tentative or Any Day activities Occasionally question dialogs may also show up if respondents have entered certain activities that the survey administrators are interested in investigating After users finish updating their Fina
18. ncy locations and involved person Data obtained from the interview were manually entered by the interviewer into a computer database This database is linked to the CHASE program so later in the schedule reporting sessions users can select without typing one of the alternatives they indicated in the interview After the up front interview respondents were trained to use the CHASE program to record their weekly activity scheduling process Figure 1 shows the main user interface This interface was built upon a day planner software component that displays schedules in a calendar format with time scale on the left Participants were asked to log into the program at least once a day for the entire week On the first Sunday night they would add activities anywhere in the calendar from Monday to Saturday that they have already thought about doing before launching the program On Monday they would enter data through the program to reflect what they had done for the day Respondents were then asked to review modify and add activities for subsequent days Tuesday to Sunday If any changes to these future activities were known at this moment they should enter the changes The same process will continue every day until Sunday Despite the efficiency demonstrated in the original CHASE survey several areas for improvement have been identified First a laptop computer needs to be placed in a household for the week long duration of the survey and field wo
19. now Don t remember don t know 2 don t remember o M Eat Sleep Wash Meal at home or at work M Eat Sleep wash Night sleep O Eat Sleep wash Snack M E at Sleep wash Wash dress pack U Household Obligation Attending to children at home O Household Obligation Cleaning Maintenance at home M Household Obligation Filing out CHASE survey U Household Obligation Meal preparation Household Obligation Other household obligations O Household Obligation Pick up drop off others O Household Obligation Pick up drop off kids i Net gt Finish Figure 4 Activity Program Selection Initial Interview Click at the yellow page to get online help Activity Time and Days Setting For sm Activity Title Activity 1 of 6 bill chool Work IF you don t feel comfortable answering this question just click Next gt and don t change anything Does this activity occur on fixed _ _ Does this activity have a fixed Ka EF day s of the week SAG start end time Yes C No C No Which Days I Monday Thursday Hour Min Start Time I Tuesday I Friday BR Hjo nz Z Wednesday Fm Saturday Ent Tine x o z G AM C PM Sunday Exit for now and continue later i i Finish Figure 5 Activity Time and Day Entry Form Initial Interview Click at the yellow page to get online help FE Shopping Major Grocery 10 items Campus Plaza on Campus and C
20. ntries on the first Sunday evening Any Day Activities Although all participants 15 individuals knew the meaning and purposes of the Any Day column see question 7 of Table 6 4 of them did not enter Any Day activities on the fist Sunday A total of 26 activities were indicated as Any Day activities Table 5 categorizes the activities entered as Any Day Lee et al 11 Responses to the Decision Tracing Dialogs During the first week of pilot study REACT collected from participants 1256 responses as to why they made certain schedule changes Changes made to basic activities e g night sleep shower regular meals were not questioned The purpose was to keep the length of each data entry session short If changes to all activities were questioned the excessive number of dialogs may make respondents impatient thus the quality of the data obtained could be affected Further analysis on these responses is pending until the target sample size is reached However initial analysis indicated a majority of cancellation of activities were due to entry errors because participants learned how to use the program on their own A step by step user manual is now shipped with the survey package intending to shorten the learning periods User Evaluation After the end of the surveying week a set of 11 questions see Table 6 were sent to the 9 participating households by email Since couples were allowed to learn how to run the program from each
21. o run the program user manual and program installation guide was subsequently sent to each participant see Appendix for the survey guidelines given to respondents After that participants would execute all the tasks on their own An exclusive phone line was dedicated for technical support The survey administrator would occasionally send out email to alert interviewees potential mistakes they might make on a specific day Emails would also be sent to remind participants if investigators detect individual households that forgot to send in data on the previous day After the survey ended a set of questions were sent by email to every participating household asking for their evaluation of the program see User Evaluation After the beta testing a version of the program for pilot study was finalized Lee et al 10 Preliminary Results from the Pilot Study The pilot study began from the second week of April 2000 and will continue until the target sample size of 50 households is reached Currently 9 households 3 singles and 6 couples totaling 15 individual participants living in two student communities around UCI campus completed the survey These two communities are located within 10 minutes of walking distance from the main campus It has to be noted that these 9 households are different from those participating in the beta testing Respondents are graduate and undergraduate students of UCI and their spouses The average age of all respondents
22. of this paper is to describe features of REACT and to present preliminary results from a field testing conducted in a campus community around University of California Irvine UCI Lee et al 3 REVIEW OF CHASE The main objectives of CHASE were to explore a household s activity agenda from which all activities are drawn and to track the entire process of when and how activities from the agenda are added deleted and subsequently modified in a week long period These were accomplished through a household interview self completing data entry of a weekly activity schedule through the CHASE program and a follow up interview Forty households from Hamilton Ontario Canada were recruited to participate in a pilot survey The program was installed on three laptop computers rotated amongst households on a weekly basis Laptops were dropped off on a Sunday evening and picked up the next Sunday evening Up front interviews were conducted on weekends before Sunday evening and lasted 1 5 to 2 hours The purpose of the interview is to obtain information on household demographics available transportation modes and residential information along with the household s activity agenda A household s activity agenda is the set of activities in which the household members participate on a weekly basis Household members were asked to describe in their own words the specific activities of each type that they perform along with their attributes such as freque
23. on of the behavioral process that un revealed travel behavior To reveal this process a new computer program REACT L has been developed to collect data for a study of the determinants of travel and activity behavior in households This data is inherently dynamic since respondents record planned activity schedules and then update these schedules on a daily basis fully defined in time and space The resultant data will facilitate the identification of fundamental inter relationship among a comprehensive range of revealed travel and activity participation variables leading toward the identification of what are the critical variables relationships and rules that govern that behavior It is believed that an internet based travel survey particularly one as rich in resultant content as REACT will significantly reduce data collection costs improve data quality and quantity and allow for continuous data collection The purpose of this paper is to describe features of REACT and to present preliminary results from a pilot study An earlier version of the program was named iCHASE and documented in Lee et al 2000 Lee et al 2 INTRODUCTION An activity schedule defined by Axhausen 1995 represents the joint choice of the time duration location mode and route for a sequence of activities drawn from a given set of aware activity needs Household activity scheduling extends the context of activity scheduling by considering the effect
24. ouseholds if models addressing new policies are to be developed Lawton 1996 also noted the inefficiency of the current data collection methods and stated We should seriously evaluate the use of more carefully chosen smaller samples using direct contact and paying for cooperation their time Data collection needs to be automated laptop etc and we need to design interactive stated response experiments that key directly from revealed data at the same collection time The Computerized Household Activity Scheduling Elicitor CHASE program developed by Doherty and Miller 2000 represents one of the efforts dedicated to advance methods for collecting data of household activity travel behavior The program was installed on laptop computers rotated among surveying households to record weekly household activity schedules It broadened the dimensions of household activity travel diaries by questioning the entire decision process from pre travel planning to post travel schedules in a week long span Despite the efficiency of the program several areas for improvement have been identified see next section A new program REACT addressing CHASEF s inefficiencies has been developed REACT allows respondents to use their own computers to enter activity diaries and upload them to a server In addition to hardware and software enhancement significant advancement is made in terms of tracing decisions involved in the scheduling process The purpose
25. r frequent locations If a desired location is not included in the provided list users can choose to enter this location by typing the address closest intersections or pointing it out in a built in GIS Figure 3 3 Household Vehicles On this form users enter the make model and year of the vehicles in the household 4 Household Members The first person in the household to run initial interview will enter the first names of all household members A copy of individual Lee et al 6 database for each member will be generated to store personal information and activity diaries 5 Member Information Every adult member in the household will enter his her personal information on this form including Gender age license to drive relation to household head education level employment status hours worked in a typical week occupation student level if the person is a student 6 Activities Information This group of questions is intended to investigate what activities this household member usually do i e at least once per month Attributes e g typical frequency duration time windows locations and involved persons of these activities will also be explored Within this group there are 5 subordinate forms a Activity Selection Figure 4 Users will first select their typical activities i e performed at least once per month from the list of activities provided see Table 1 If an activity is not in the list they can always
26. rkers were required to deliver laptops and carry out the up front interviews In order to expand sample sizes and reduce costs such an approach would need to be augmented through the use of home computers and or remote access to the program The interview also needs to be computerized Second although the calendar like interface greatly accelerated the process of entering activity schedules it is not known if such an interface biased the decision process It is reasonable to suspect that if a time table is presented to the respondents they might be tempted to fill up the gaps by inserting plans they wouldn t have made under normal circumstances To be more specific when they see their Lee et al 4 activities laid out on the time dimension it may encourage people to arranged things better resulting in more scheduling steps which they would not otherwise have made Third CHASE does not allow the respondents to leave certain activity attributes undetermined when preplanning except in the case of mode and travel time information In reality people s plans may often remain only partially elaborated Modifying the program to allow for this would enhance the opportunities for understanding the nature of activity scheduling process REACT PROGRAM DESIGN System Architecture The most urgent improvement over the current CHASE approach is to reduce the cost and human resource associated with using laptops as survey instruments One poten
27. tabase linked to Weekly Activity Diaries Thus participants can enter details e g activity titles locations and other people involved of an activity by selecting from what they have entered in the interview Typing is not necessary for most of the entry items Initial Interview When users launch the program for the first time they will be guided to complete the Initial Interview Questions asked in the initial interview are categorized into the following 7 groups Question groups 1 to 4 are related to the entire household Only the first person doing the interview need to answer these questions Next member to do the interview will begin from group 5 1 Household Information Users will enter their home address and answer the following questions a Do you own or rent the household b What s the type of your household e g condo apartment and townhouse c How long have you stayed at the current residence d What s your annual household income 2 Frequently Visited Locations This form Figure 2 is intended to collect a list of locations the respondent household visit at least once per month This list will be available to respondents when they need to indicate activity locations in the Weekly Activity Diaries In addition the list enables researchers to grasp respondents action space and to gain better understanding of their spatial behavior Users will select from the provided list of local activity locations and built thei
28. termined by the number of days you took part in the study Included in this envelop are a program CD a brief program and installation guide a step by step guide to tell you how to run the program on Monday and two Consent to Act as a Human Research Subject forms Please read this form before you sign and proceed with the survey The consent forms are required by Institutional Review Board at UCI s Office of Research Administration If you are single you only need to sign one If you and your spouse will both participate in the survey both of you should sign Please try to install the program before this coming Sunday and run it after installation to begin the Initial Interview For instruction on how to install the program please see the REACT Installation Guide If you encounter problem installing the program please contact us as soon as possible If you have any questions and or problems with the REACT program please feel free to contact Ming Lee at 949 824 6571 or 949 233 6094 cellular phone or send email to mingshel uci edu Thank you very much for taking part in the REACT survey
29. tial way of resolving this dilemma is to conduct the survey via Internet Currently there are two distinct remote computing systems that can operate across the Internet Web based and Fat Client applications A Web based application performs most of the processing within a remote Web server Such an application is sometimes referred to as a thin client Any computer with a Web browser and Internet connection can access the application and interact with its data records residing on the server Although a Web application s ability to reach a broad spectrum of users makes it an idea system to implement a survey application there are several limitations First interactivity between an application and users is slow and limited Data need to be frequently transferred back and forth between a client and server with users waiting for the next batch of data to come in prior to proceeding with the survey Second the stability of a Web applications is subject to the amount of network traffic Finally users have to remain on lined for the entire survey session Consequently survey participants have to absorb the connection cost for the time they run the survey A remote computing system that performs most of the data processing within its client is usually referred to as a fat client system Any stand alone program that runs within a client computer and has the ability to connect to a remote server is by definition a fat client application The fat client
30. ticipants wanted for research project The Department of Civil and Environmental Engineering of UCI is conducting research aiming at collecting information of what activities people typically do on a weekly basis This information is expected to help us to estimate the amount of daily traffic on streets We are looking for single adults over 18 years of age or married couples with or without children who have a PC running Windows 95 98 or NT at home with Internet connection Participants will install a computer program and enter their weekly activities in this program Each participant will typically spend approximately 20 to 30 minutes each day for 9 days from a Sunday to the next Sunday plus one day before the first Sunday Participants will be financially compensated for participation in the study single persons will receive 25 dollars married couples will receive 50 dollars If participants decide to withdraw from participation before the data collection is completed partial compensation i e proportional to the number of days participated will be provided Please contact Ming Lee at mingshel uci edu or 949 824 6571 for detailed information or visit survey administrative Web site at www its uci edu react GUIDELINES DISTRIBUTED WITH THE REACT SURVEY PACKAGE Thank you for taking part in the REACT survey Welcome to the REACT survey The survey has two parts Initial Interview and Weekly Activity Diary The recording of your Weekly
31. ut REACT survey minutes Church and other religious Medical care events Personal services Hair nails Volunteer work Professional services dry clean auto repair Kids Activities Tag along with parents Banking ATM Play socializing Post office Shipping Homework Library With babysitters Rental video return Practice lessons Gas station Video rental store Major Grocery 10 items Minor Grocery lt 10 items Housewares clothing personal l items Drug Store Mostly browsing Convenience store Table 1 Pre defined Activity List Activity Manipulation Question ID Questions Triggered An intended activity was 1 Why did you cancel this activity canceled during the day An activity was executed at a 2 Why is the time different from what time different from the you entered before intended An intended activity is Why is the day different from what postponed you entered before An activity was executed at a 4 Why is the location different from location different from the what you entered before intended A specific activity is added to 5 Why do you want to do this activity on the Weekly calendar this day An activity not known before the day was executed during the day The start or end time of an executed activity was originally left as I don t know The location of an executed You did not enter the location for the activity was originally left as activity before How did you decide to I don t know
32. vity record all the triggered questions will show up in turn Each of these questions will appear in a dialog box like Figure 13 It is important to note that interviewees are warned that they should not force themselves to come up with a reason In case the decision was made without much reasoning they can always select an option called No specific reasons Depending on the answer selected subsequent dialogs can be triggered to trace other information that may be relevant to the decision First the timing when the decision was made may be asked with the dialog box in Figure 14 This dialog may be followed by a simple question Do you often make such a decision in a similar situation Finally the last dialog Figure 15 may be used to question if respondents ever thought about anything as the benefit of the manipulation Lee et al 9 Send Data Utility When all the adult members in a household finish recording their activity diaries the last person will establish Internet connection then activate the Send Data utility of REACT Although the structure of REACT s central database is capable of being served from a remote database server currently the databases reside on the clients end to achieve privacy and efficiency with a minimum deployment cost When Send Data is activated the program will compress the database of the entire household with encryption The data package will be labeled with the household s ID and current date and sent to
33. y include appointments and regulars such as work regular exercise or pickup and drop off children You should simply enter the activities on the days they usually get dane Do NOT intentionally think about how you can better arrange your schedule if you don t usually do so at the beginning of a week Next gt eee MIE ri NaXx D p7 p0 YXSESRnXBX gt yyZ Figure 10 Data Entry Guidelines Given by the Automatic Slide Show Activity Information EnerAtv y don t know 2 don t remember Monday mr C People outside my household Figure 11 Activity Information Form Mon 3 27 00 Figure 12 Adding Activities and Travel Events to Final Schedule You can check more than one from the followings But PLEASE DO NOT FORCE YOURSELF TO COME UP WITH A REASON NOW If you don t think that you ever thought about anything at the time feel free to choose No specific reasons j r r r r lm m m If you select an answer that is followed by an option box please select one from the box to complete this answer If you select Other reaons please describe briefly the reason in the following text box Figure 13 Decision Tracing Dialog s s oi La 2 2 Figure 14 Decision Timing Tracing Dialog Have you thought about ln l l k l lm po Figure 15 Decision Benefit Tracing Dialog APPENDIX RECRUITING ADVERTISEMENT Par
34. you saw it for the first time If not could you tell us what we can do to improve it 11 Do you have any comments on any aspect of the survey research process NA NA NA One respondent did not answer questions 9 and 10 Table 6 Results of User Evaluation Activity Scheduler Exit Add Modify Delete Help Thursday Saturday 03 00 PM oe 04 45 PM i i Heva 2 a 3 8 ajaja a an mea Figure 1 Main Interface of CHASE Initial Interview Click at the yellow page to get online help Frequently Visited Locations 1 Please review the list of locations on the left and select the ones you visit at least Help once per month and move them to the list on the right by pressing the right arrow button 2 Be sure to move Home to your Frequent Locations Locations in the area Frequent Locations Location Description Ci _ Frequently Visited Location Home Alton Square on Alton and Jeffrey Ben l vi and Jerry Z Pizza OCTFCU Ralphs Arbor Village Center on Walnut and J effrey Irvi T aiko Weinerschnitz Campus Plaza on Campus and California lrvi Albertson s Culver Plaza on Irvine Center Dr and Irvi Culver 99 Ranch Petco Three D Figure 2 Entry Form for Frequently Visited Locations ia Right Location Figure 3 Built in GIS for New Location Entry Initial Interview Click at the yellow page to get online help M Don t k
Download Pdf Manuals
Related Search