National Travel Survey: Data Extract User Guide
Contents
1. The weighted number of individuals also needs to be known From the menu select Data gt Weight cases Check Do not weight cases Select OK First turn the weighting off Next we must ensure that we are only counting individuals who have been asked their From the menu select Data gt Select cases Select If condition Edit the definition to select individuals answering the question The box should read SurveyYear 2012 amp ReSNDN_NA_BO1ID 2 amp ResNDN_DNA_B01ID 2 amp ResNDN_DEAD_B01ID 2 Select OK reasons for not learning to drive eS SS SS SS SSS NTS data extract 2002 2012 user guide Page 30 of 32 From the menu select Analyze gt Compare Means gt Means To count the weighted number of individuals we need to count W3 Move W3 across to Dependent List Select Age_BO4ID as the layer 1 variable by moving it across to the independent list selection Select Options to determine the type of statistics to perform Move Sum across to the Cell statistics box Remove all other statistics Select continue Select OK The tabulation can now be specified The following SPSS output should be produced Report Age of person banded Weighted age Band D All ages interview 9 categories sample 17 20 years 447 1548920 29 years 696 8862122 39 years 531 8084065 49 years 449 5771186 59 years 470 9010953 60 years2. PSU Primary Sampling Units PSU are used to provide a maA list of areas from which to select a LDJ Li Drog sample of addresses sourmey from In this case l Household each PSU is a Trip postcode sector Key identifier variables To enable the linking of data from each level of the dataset hierarchy a set of identifier variables are attached to each data file The combination of identifier variables that uniquely identifies a record at each level of the database are PSU PSUID Household HouseholdIiD Vehicle VehiclelD Individual IndividuallD LDJ LDJID Day DaylD Trip TripID Stage StagelD Ticket IndTicketID To link levels together the identifier variables from the highest analysis level should match those at the lower level These identifier variables are new unique IDs which allow for linkage between data levels without requiring linkage on multiple variables as in previous versions of the datasets held by the UK Data Service SESEESESSSSoEeE EHHR TEE SS SS SS EEE SSS Same NTS data extract 2002 2012 user guide Page 5 of 32 Example To link an individual to the trips they made merge the individual and trip level files and ensure that the individual level identifier IndividuallD matches on both files Individual IndividuallD Trip IndividuallD If merging datasets together users should remove unwanted variables to reduce the file sizes and length of time it t
3. 1161 584938 Total 3757 912662 To calculate all percentages as in NTS0203 divide the weighted number of individuals for each response and age group by the relevant total number of individuals e g Percentage of individuals aged 17 20 with the reason Family or friends can drive me when necessary given 132 9655693 447 1548920 Repeat the above step with WO no weight to get the unweighted number of individuals NTS data extract 2002 2012 user guide Page 31 of 32 x 100 30 The following SPSS output should be produced Report Sum Age of person banded Unweighted age Band D All ages interview 9 categories sample 17 20 years 21 29 years 30 39 years 40 49 years 50 59 years 60 years Total SSE SSE EE Samay NTS data extract 2002 2012 user guide Page 32 of 32
4. Statistics gt Frequencies Move Survey Year across to the Column box Select OK The tabulation can now be specified The following output should be produced Survey year actual year Cumulative Frequency Percent Valid Percent Percent 8 2 8 2 8 2 17494 19578 19302 20103 19794 19940 19253 19918 19210 18339 19547 212477 NTS data extract 2002 2012 user guide Page 26 of 32 The stage distances can now be divided by the individual sample the frequency column It should be noted that the stage distances are for a single week so therefore need to be multiplied by 52 14 number of weeks in a year 66278 50 e g Trips per male in 2002 x 52 14 17 494 197 54 As in Annex A repeat the above step using W1 to get the unweighted number of individuals This needs to be known to ensure that the base sample sizes are sufficient Select the Individual sav file From the menu select Data gt Weight cases Check Weight cases by and move across W1 unweighted diary weight as the frequency variable Select OK The tabulation can now be specified From the menu select Analyze gt Descriptive Statistics gt Frequencies Move Survey Year across to the Column box Select OK The following SPSS output should be produced eq SS SS SS ae NTS data extract 2002 2012 user guide Page 27 of 32 Survey year actual year Cumulative Frequency Percen
5. across to the column specification box Select Cells and in the Non integer weights section select No adjustments Select OK The following SPSS output should be produced for each response variable there are 16 possible responses A P LLL SS ES Saas NTS data extract 2002 2012 user guide Page 29 of 32 Reason individual does not drive Family or friends can drive me when necessary Age of person banded age Band D All ages 9 categories Survey year actual year Crosstabulation Count Age of person banded age Band D All ages 9 categories Survey year actual year 0 4 years 5 10 years 11 16 years 17 20 years 21 29 years 30 39 years 40 49 years 50 59 years 60 years Reason individual does a 000 000 000 132 966 166 050 133 267 111 638 134 147 417 959 1096 027 not drive Family or friends can drive me No when necessary 1348 569 1454 680 1525 500 910 225 2419 969 2700 392 3063 582 2520 721 4345 085 20288 725 Total 1348 569 1454 680 1525 500 1043 191 2586 019 2833 659 3175 221 2654 868 4763 045 21384 752 Reason individual does Yes not drive Family or friends can drive me No when necessary Total 1348 569 1454 680 1525 500 1043 191 2586 019 2833 659 3175 221 2654 868 4763 045 21384 752 000 000 000 132 966 166 050 133 267 111 638 134 147 417 959 1096 027 1348 569 1454 680 1525 500 910 225 2419 969 2700 392 3063 582 2520 721 4345 085 20288 725
6. can be treated in this way A doctor s round could therefore consist of one trip to the first patient one series of calls trip to the other patients and one trip from the last call back to the surgery or home A full list of definitions is available in the 2012 Notes and Definitions document available at https www gov uk government uploads system uploads attachment_data file 226817 nts201 2 notes pdf a T NTS data extract 2002 2012 user guide Page 8 of 32 Weighting Short walk weighting Because trips of less than one mile in distance are recorded only on the seventh day of the travel week these trips must be weighted by a factor of seven when analysed Also for consistency with earlier surveys series of calls trips are excluded from analysis of stage and trip counts and time Therefore one of the following short walk weights must be applied to any tabulations using trip or stage counts distance or time SSXSC Number of stages to be counted grossed for short walks and excluding series of calls trips SD Stage distance travelled grossed for short walks STTXSC Stage travelling time grossed for short walks and excluding series of calls trips JJXSC Number of trips to be counted grossed for short walks and excluding series of calls trips JD Trip distance travelled grossed for short walks JOTXSC Overall trip time includes travelling and waiting time grossed for
7. person per year by gender and survey year Load the journey level SPSS data file Trip sav Load the individual level SPSS data file Individual sav Before performing the analysis you should ensure that both files are sorted based on the key identifier variables at each level Select the Trip file From the menu select Data gt Sort Cases Highlight TripID and move over to the Sort by list Ensure that ascending is checked The Sort by list should contain TripID A Select OK Select the Individual file From the menu select Data gt Sort Cases Highlighting IndividuallD and move over to the Sort by list Ensure that ascending is checked The Sort by list should contain IndividuallD A Select OK LL SSSSSSSSSSEaEaE a ae NTS data extract 2002 2012 user guide Page 16 of 32 To disaggregate the data by the gender of the individual making the trip individual level data needs to be attached to the trip file Select the Trip file From the menu select Data gt Merge files gt Add variables Select An open dataset and then select file Individual sav Select continue Data should be linked using the identifier variable at individual level IndividuallD Check Match cases on key variables and Cases are sorted in order of key variables in both datasets Check Non active dataset is keyed table From the excluded variables list highlight Individual
8. sav Before performing the analysis you should ensure that both files are sorted based on the key identifier variables at each level Select the Stage file From the menu select Data gt Sort Cases Highlight StagelID and move over to the Sort by list Ensure that ascending is checked Select OK Select the Individual file From the menu select Data gt Sort Cases Highlighting IndividuallD and move over to the Sort by list Ensure that ascending is checked Select OK SSE eS SS SS ee SSS NTS data extract 2002 2012 user guide Page 23 of 32 Before producing the tabulation of stage distance by survey year and mode the weighting must be specified Select the Stage file From the menu select Data gt Weight cases Check Weight cases by and move across W5 stage weight as the frequency variable Select OK The tabulation can now be specified From the menu select Analyze gt Compare Means gt Means To gross up short walks we must use variable SD as the dependent list variable Move SD across to Dependent List Select SurveyYear as the layer 1 variable by moving it across to the independent list selection Select Next to choose the 2 layer variable Select Stage Mode as the layer 2 variable by moving StageMode_BO01ID across to the independent list selection Select Options to determine the type of statistics to perform Move Sum across to the Cell statistics box
9. travel diary Short walk weights account for this and should be applied when producing analysis see Weighting section below For detailed walking trip analysis it may be necessary to limit the base sample to day 7 only data e g NTS9909 SSE SS SS SS Same NTS data extract 2002 2012 user guide Page 7 of 32 Key Definitions Trip A trip is defined as a one way course of travel having a single main purpose e g a walk to school or a trip to work without any break in travel Usually the respondent is clear what the single main purpose of a particular trip is However sometimes people go out for a number of reasons or go out for one main reason but carry out a number of different activities perhaps at different places Complex travel like this is broken into separate trips so that the data can be analysed Where a stop is entirely secondary to the main purpose such as a stop to buy a newspaper on the way to work the stop is disregarded Stage A trip consists of one or more stages A new stage is defined when there is a change in the form of transport or when there is a change of vehicle requiring a separate ticket In order to reduce the burden on respondents travel involving a number of stops for the same main purpose and using the same form of transport are treated as one continuous series of calls trip from the first such call to the last one Only shopping and in course of work travel
10. 32 Fully co operating households Diary sample Interview data travel data Partially co operating households Interview data only It is important to use the correct sample for all analysis The weighting methodology produces weights at the household LDJ and trip level The household weights apply to all individuals and vehicles within the household and they have therefore been attached to the individual and vehicle files for ease of use Similarly the trip level weights apply to all stages within trips and have therefore been attached to the stage level records for ease of use The variable OutCom_BO2ID 1 fully productive O partially productive is included on the household file to enable users to select the diary sample This variable can be merged into other files if necessary e g individual file Weighting matrices The following weighting matrices are available and should be used on the appropriate level data for the analysis being produced W1 Unweighted diary sample this gives unweighted results for the diary sample only This is equivalent to the results produced before the weighting strategy was introduced and can be used to generate unweighted sample sizes for analysis of the diary sample It is effectively the same as the OutCom_BO2ID variable mentioned above formerly named the status variable W2 Diary sample household weight adjusts for non response apply to all analys
11. 36 Unweighted diary short walks grossed up amp 174 261 169 384 166 421 150 833 151 978 172 463 167 939 349 251 396 761 391 871 408 775 399 973 384 402 371 642 390 741 370 279 345 030 367 019 excluding series of calls Unweighted interview 7 x 2 2 A Weighted diary 135 017 139 423 132 494 125 838 128 346 144 406 143 953 314 728 348 024 341 321 361 449 352 392 340 520 335 877 335 727 324 118 305 012 326 477 Weighted diary short walks grossed up amp 177 225 180 055 172 061 160 462 162 999 181 506 179 767 378 777 419 996 412 966 437 211 430 032 407 550 401 690 408 304 392 572 370 194 395 017 excluding series of calls Weighted interview 1 Data for 1995 to 2001 is based on calendars years Data for 2002 to 2012 is based on survey years which run mid January to mid January https www gov uk government uploads system uploads attachment_data file 225736 nts201 2 sample xls Note Between 1995 and 2001 published NTS analysis is based on the calendar year in which a household started completing their travel diary Analysis from 2002 onwards is based on the survey year mid January to mid January in which the household was sampled LL SSSSSSEEEEaEE SES ee ee ae NTS data extract 2002 2012 user guide Page 15 of 32 Annex A Example tabulation 1 NTS data users should replicate analysis produced by DfT to ensure that they understand the methodology of weighting and the grossing up of short walks Trips per
12. 65 21 835 20 839 19 988 21 243 Weighted diary 7 675 7 704 7 487 6 986 7 109 8 114 8 132 17 494 19 578 19 302 20 103 19 794 19 940 19 253 19 918 19 210 18 339 19 547 Weighted interview 8 348 8 405 8 258 7 857 7 945 8 942 8 811 20 789 21 795 21 369 22 539 22 098 21 959 21 219 21 706 20 819 20 057 21 385 Child lt 16 yrs sample Unweighted diary 1 744 1 666 1 650 1 433 1 466 1 731 1 659 3 413 4 178 4 129 4 150 3 938 3 971 3 804 4 042 3 865 3 540 3 661 Unweighted interview 1 945 1 869 1 917 1 708 1 703 1 971 1 860 4 398 4 702 4 690 4 776 4 497 4 408 4 260 4 414 4 193 3 921 4 115 Weighted diary 1 602 1 634 1 547 1 447 1 467 1 666 1 624 3 523 3 900 3 815 3 963 3 848 3 759 3 684 3 790 3 660 3 524 3 718 Weighted interview 1 741 1 782 1 704 1 626 1 631 1 835 1 764 4 181 4 336 4 222 4 443 4 296 4 141 4 057 4 118 3 962 3 852 4 060 Adult 16 sample Unweighted diary 5 979 5 999 5 823 5 409 5 504 6 325 6 319 13 473 15 289 15 070 15 754 15 552 15 764 15 179 15 872 15 207 14 529 15 493 Unweighted interview 6 576 6 635 6 535 6 237 6 297 7 083 6 973 16 429 17 288 16 898 17 926 17 644 17 523 16 905 17 421 16 646 16 067 17 128 Weighted diary 6 074 6 069 5 940 5 539 5 640 6 447 6 508 13 971 15 678 15 487 16 141 15 945 16 181 15 569 16 128 15 550 14 815 15 829 Weighted interview 6 606 6 623 6 553 6 231 6 315 7 107 7 047 16 608 17 459 17 147 18 097 17 801 17 818 17 162 17 587 16 857 16 206 17 325 Motor vehicle sample Unweighted diary 3 296 3 301 3 238 3 121 3 217 3 772 3 707 8 195
13. 9 264 9 065 9 847 9 758 9 898 9 511 9 985 9 750 9 192 9 702 Unweighted interiew 3 642 3 691 3 653 3 608 3 681 4 240 4 081 9 954 10 452 10 190 11 228 11 118 10 975 10 588 10 911 10 675 10 136 10 678 Weighted diary 3 339 3 374 3 313 3 201 3 270 3 843 3 780 8 391 9 408 9 261 10 059 9 875 10 005 9 655 9 956 9 790 9 214 9 664 Weighted interview 3 629 3 694 3 663 3 601 3 664 4 237 4 090 9 959 10 465 10 270 11 264 11 028 10 995 10 636 10 841 10 594 10 079 10 569 Trip sample Unweighted diary 126 088 124 748 122 397 112 867 114 501 130 179 129 998 278 916 314 845 310 065 322 500 312 347 305 077 295 791 306 743 290 803 272 208 289 573 Unweighted diary short walks grossed up amp 156 729 153 366 150 403 137 851 138 233 155 644 155 789 326 495 369 664 363 692 379 315 369 072 353 191 341 529 358 500 337 402 316 494 335 288 excluding series of calls Unweighted interiew R 5 x 2 a 5 x a 2 5 7 Weighted diary 129 356 133 896 127 242 120 996 123 182 137 689 139 240 302 796 333 833 326 869 345 996 336 802 324 882 319 902 320 257 308 431 291 072 310 732 Weighted diary short walks grossed up amp 159 163 162 418 154 703 146 148 147 271 162 950 165 284 351 123 388 152 379 887 402 356 393 812 371 823 366 326 371 473 354 900 337 029 357 616 excluding series of calls Weighted interview Stage sample Unweighted diary 131 548 129 690 127 273 117 269 119 072 136 324 134 036 289 048 327 230 322 602 335 940 326 076 318 996 309 722 320 858 304 942 284 718 303 4
14. Remove all other statistics Select continue Select OK SEES SS SS EE E Samay NTS data extract 2002 2012 user guide Page 24 of 32 The following output should be produced for all years down to 2012 Stage Stage mode oftravel as distance recorded in diary 18 grossing up categori short walks 66278 50 Bicycle 12132 08 Private hire bus 41637 52 Car 1845539 78 Motorcycle scooter moped Van lorry 92132 81 Other private 6914 02 London stage bus 18742 35 Other stage bus 75319 03 Coach Express bus 10923 44 Excursion Tour bus 9005 67 London Underground 27065 54 Surface Rail 138536 02 Light rail 2536 75 Air 14752 25 Taxi 13266 63 Minicab 6594 07 Other public 1073 24 Total 2394079 50 11629 80 LEE SS T A NTS data extract 2002 2012 user guide Page 25 of 32 These are the weighted and grossed up number of stage distances per week for each survey year mode combination For each mode survey year the number of individuals also needs to be known Note it is standard NTS methodology to include all respondents in the denominator not just those making a trip Select the Individual sav file From the menu select Data gt Weight cases Check Weight cases by and move across W2 household weight as the frequency variable Select OK Before producing the tabulation of mode by survey year the weighting must be specified From the menu select Analyze gt Descriptive
15. UK Data Archive SN 5340 National Travel Survey 2002 2012 AEN National Travel Survey Department Data Extract User Guide 2002 2012 for Transport Contents Data Update 2 Key Issues 3 Survey Overview 4 Using NTS Data 7 Key Definitions 8 Weighting 9 Revisions 14 Sample Sizes 15 Annex A example tabulation 16 Annex B example tabulation 23 Annex C example tabulation Multi coded 28 data E SS SS SS EEE ae NTS data extract 2002 2012 user guide Page 1 of 32 Data Update 2002 2012 datasets supersede the 2002 2010 SN 5340 files The National Travel Survey NTS team in DfT recently completed the redevelopment of the database in which all NTS data is held All historic data 2002 2007 was copied from a Quantum Quanvert system into a new SQL database The NTS data from 2008 to 2012 has been fully processed in SQL and this has over written previous data for years 2008 to 2010 which was previously processed in Quantum and supplied to the UK Data Service In addition as part of the database redevelopment all processing methods were reviewed including imputation of missing data for some variables Improvements and corrections have been made which has resulted in some minor revisions to all data published from 2002 See Revision section of this document for details of the main changes made which have affected time series results These datasets contain the last set of NTS data covering residents of Great Britain A
16. akes to merge See examples in Annex A and B for detailed instructions of how to merge data files T A NTS data extract 2002 2012 user guide Page 6 of 32 Using NTS Data Distance The length of any trip stage is the distance actually covered as reported by the traveller and not the distance as the crow flies Distance is measured in miles previous versions of the datasets held by the UK Data Service used tenth of mile Distance by mode Average distance travelled results are calculated based on the stage distance for each mode and are presented as average distance travelled by mode see table NTS0305 Trip rates are usually presented as main mode i e based on the mode that is the longest stage distance see table NTS0303 Time Unit minutes Denominator It is standard NTS methodology to include all respondents in the denominator not just those making trips For example children should be included when calculating driver trips rates Number of records There are less records in 2011 due to the GPS experiment which was conducted using a sub sample and data collected was not incorporated into the main database Mode Taxi minicab When using main mode or mode variables the split between taxi and minicab should not be used as it is not accurate Results should be presented combined like published tables e g NTS0303 Mode Walking Short walks those under one mile are only collected on day 7 of the
17. ince 2002 For advice on using and interpreting the NTS data please contact national travelsurvey dft gsi gov uk SSE SS SS Same NTS data extract 2002 2012 user guide Page 3 of 32 Survey Overview Survey background The NTS is carried out in order to monitor long term changes in personal travel and provide a better understanding of the use of transport facilities made by different sectors of the population Individuals in sampled households are interviewed face to face to collect personal information such as age gender working status car access and driving licence holding They are also asked to complete a seven day travel diary and provide details of trips undertaken including purpose method of travel time of day and trip length The first NTS survey was commissioned by the Ministry of Transport in 1965 with further periodic surveys being carried out in 1972 73 1975 76 1978 79 and 1985 86 There area number of differences between the current survey and surveys up to 1978 79 These are detailed in Chapter 7 of the National Travel Survey 2008 Technical Report which is available on the archived NTS website at http webarchive nationalarchives gov uk 20091203140650 http www dft gov uk pgr statistic s datatablespublications personal methodology ntstechreports ntstechrep2008 Conducting the survey Prior to the interviewer s first call letters are sent out to the sampled addresses These introduce the survey and explain that a
18. is of the diary sample at household individual and vehicle level W3 Interview sample household weight apply to all analysis of the interview sample at household individual and vehicle level W4 LDJ weight incorporating household weight apply to all analysis at LDJ level LS www SSS SS SS A NTS data extract 2002 2012 user guide Page 12 of 32 W4xhh LDJ weight excluding household weight apply only if also applying a household weight i e W2 or W3 W5 Trip stage weight apply to all analysis of trip and stage data W5xhh Trip stage weight excluding household weight apply only if also applying a household weight i e W2 or W3 No weighting matrix or WO If no weighting matrix is applied this gives unweighted results for the interview sample e For most analyses at household individual and vehicle level w3 should be applied e For most analyses of travel patterns w5 should be applied to trip stage data and w2 should be applied at the individual level in order to calculate rates In addition the short walk weights should be applied see Short walk weighting above Examples of applying weights e To generate trip rates apply w5 to trip data and apply w2 to individual data i e Diary sample e To calculate household car ownership apply w3 to the household data Interview sample e To calculate the proportion of driving licence holders apply w3 to the individual data Interview sample e To dete
19. lD and move to the key variables box Select OK A message will appear Warning Keyed match will fail if data are not sorted in ascending order of key variables This can be ignored as we have already ensured the correct sorting order The individual characteristics of the person making each trip will now be attached to the journey data Before producing the tabulation of gender by survey year the weighting must be specified From the menu select Data gt Weight cases Check Weight cases by and move across W5 trip weight as the frequency variable Select OK SEH SS E E Saas NTS data extract 2002 2012 user guide Page 17 of 32 The tabulation can now be specified From the menu select Analyze gt Compare Means gt Means To gross up short walks we must use variable JJXSC as the dependent list variable Move JJXSC across to Dependent List Select Gender as the layer 1 variable by moving Sex_B01ID across to the independent list selection Select Next to choose the 2 layer variable Select Survey Year as the layer 2 variable by moving SurveyYear across to the independent list selection Select Options to determine the type of statistics to perform Move Sum across to the Cell statistics box Remove all other statistics Select continue Select OK A a NTS data extract 2002 2012 user guide Page 18 of 32 The following SPSS output sh
20. ll data from 2013 onwards will cover residents of England only as the survey coverage changed All variable names have changed The Table Structures spreadsheet includes a mapping of old to new variable names This change also affects the variables used for merging files together see section on Key identifier variables below The data file containing all trip records is now called Trip sav in the 2002 2010 version it was named Journey sav A significant number of extra variables have been supplied to the UK Data Service for the End User Licence dataset Further additional variables are available to users via Special Licence and the Secure Access Portal The Table Structures spreadsheet indicates which variables are available and in which survey years in this document Full represents DfT s own version of the database E TEE Se SSS SS SSE SS Same NTS data extract 2002 2012 user guide Page 2 of 32 Key issues to consider when analysing NTS data e The NTS is primarily designed to measure long term trends in travel in Great Britain and is not suitable for monitoring short term trends e Careful attention should be paid to sample sizes for all analyses See section on Sampling errors e Appropriate short walk weights should be applied to all trip stage level analysis to account for the fact that short walks are only recorded on the seventh day of the travel week See section on Short walk weighting e In addition
21. n interviewer will call The interviewer then arranges an appointment for a placement interview During this visit the interviewer obtains information on the household each household member and on all vehicles to which the household has access via a computer assisted personal interview CAPI The procedures for the seven day travel diary record are then explained Each household is given a randomly assigned start date for the seven day travel diary The placement call is generally followed by a reminder call just before the start of the travel week to remind the household to begin their travel records and by a midweek call during the travel week to check that records are being completed correctly Within six days of the end of the travel week the interviewer will make a pick up call to collect the travel records and to check the information recorded with the informants Data types The data is hierarchical in nature and consists of several record types or levels Most significantly data is held about households vehicles individuals trips and stages Data from each level are linked to make cross level analysis possible NTS methodology determines that vehicles belong to households and not individuals Vehicle data therefore can only be analysed by household or vehicle attributes SSS SSS SSS SSS SSS EEE re NTS data extract 2002 2012 user guide Page 4 of 32 Levels in the NTS database PSU Primary Sampling Unit
22. n of missing NA and does not apply DNA cases has also been made consistent over all years Imputation indicators The NTS now includes imputation indicators in the data files for Household Vehicle Individual Trip and Stage for some key variables from 2008 to 2012 These indicators inform users if the value was originally missing and hence imputed S SSSSSSSSseo7IE SSS EE ae NTS data extract 2002 2012 user guide Page 14 of 32 Weighted and unweighted sample sizes 1995 2012 Sample numbers on which analyses are based Great Britain 1995 to 2012 Number 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Household sample Unweighted diary 3 211 3 210 3 139 2 935 3 020 3 435 3 469 7 437 8 258 8 122 8 430 8 297 8 431 8 094 8 384 8 097 7 741 8 201 Unweighted interiew 3 491 3 505 3 465 3 300 3 376 3 783 3 760 8 849 9 196 8 991 9 453 9 261 9 278 8 924 9 128 8 775 8 461 8 972 Weighted diary 3 209 3 211 3 139 2 938 3 018 3 431 3 472 7 437 8 258 8 122 8 430 8 297 8 431 8 094 8 384 8 097 7 741 8 201 Weighted interiew 3 490 3 506 3 463 3 303 3 374 3 782 3 761 8 849 9 196 8 991 9 453 9 261 9 278 8 924 9 128 8 775 8 461 8 972 Individual sample Unweighted diary 7 723 7 665 7 473 6 842 6 970 8 056 7 978 16 886 19 467 19 199 19 904 19 490 19 735 18 983 19 914 19 072 18 069 19 154 Unweighted interiew 8 521 8 504 8 452 7 945 8 000 9 054 8 833 20 827 21 990 21 588 22 702 22 141 21 931 21 1
23. ng W1 to get the unweighted number of individuals This needs to be known to ensure that the base sample sizes are sufficient Select the Individual sav file Before producing the tabulation of gender by survey year the weighting must be specified From the menu select Data gt Weight cases Check Weight cases by and move across W1 unweighted diary weight as the frequency variable Select OK The tabulation can now be specified From the menu select Analyze gt Descriptive Statistics gt Crosstabs Move Survey Year across to the row specification box Move Sex_B01ID across to the column specification box Select OK The following SPSS output should be produced E NTS data extract 2002 2012 user guide Page 22 of 32 Survey year actual year Sex of person Crosstabulation Count Survey year actual year 2002 8085 000 8801 000 16886 000 2003 9360 000 10107 000 19467 000 2004 9219 000 9980 000 19199 000 2005 9560 000 10344 000 19904 000 2006 9396 000 10094 000 19490 000 2007 9490 000 10245 000 19735 000 2008 9127 000 9856 000 18983 000 2009 9583 000 10331 000 19914 000 2010 9232 000 9840 000 19072 000 2011 8718 000 9351 000 18069 000 2012 9244 000 9910 000 19154 000 101014 000 108859 000 209873 000 Annex B Example tabulation 2 Miles per person per year by mode at stage level Load the stage level data file Stage sav Load the individual level data file Individual
24. nged imputation methodologies for key variables e g trip mode and purpose the trip data processed in SQL for 2008 2010 needed to be re weighted In most cases the overall results have changed very little from those processed in the old database Geographic data A full recode of all geographic variables at all levels was conducted so that there was a consistent time series available from 2002 2012 at household level This included creating a new Area Type variable which now correctly allocates households in South Yorkshire to the Metropolitan built up areas category This recoding also now makes the allocation of Scottish postcodes to an area type prior to 2007 consistent with 2007 onwards More information on Area type can be found in the 2012 Notes amp Definitions document All published results tables by region and area type are now all based on household geographies rather than at the PSU level so there are some minor differences Ticket data The data on tickets used and the associated costs for each stage have been re processed in a consistent way for all years from 2002 These data are now available in a separate ticket file previously they were within the individual data file Corrections Whilst working on the database redevelopment project some errors were found in the processing of some NTS data in Quantum Quanvert These errors have been corrected in the new SQL database and therefore the 2002 2012 raw datasets The allocatio
25. o select the correct weights for each analysis Initial results should be checked against published data to ensure weights are being applied correctly There are two samples which can be used for analysis Diary sample Analysis of travel data is based on the diary sample This comprises all fully co operating households defined as households for which the following information is available a household interview an individual interview for each household member a seven day travel diary for each individual and where applicable at least one completed vehicle section Weights were produced to adjust for non response and at the trip level they were also produced for drop off in recording observed during the seven day travel week Interview sample Analyses at household individual and vehicle level are based on the interview sample This sample comprises all fully co operating households included in the diary sample together with some additional partially co operating households Generally these partially co operating households had co operated fully with the various interviews but not all household members had completed the travel diary Prior to the introduction of the weighting methodology data from partially co operating households was not included in NTS analyses but this expanded dataset is now used for analyses which do not require data from the seven day travel record a NTS data extract 2002 2012 user guide Page 11 of
26. ould be produced Report Sum Number of trips grossing up short walks and excluding series of calls 170087 188735 183612 194585 188663 177980 175774 177365 169684 161512 172216 1960213 Female 181036 199417 196275 207770 205150 193844 190552 194107 185215 175516 185400 2114284 351123 388152 379887 402356 393812 371823 366326 371473 354900 337029 357616 4074497 These are the weighted and grossed up number of trips per week for each gender survey year combination NTS data extract 2002 2012 user guide Page 19 of 32 For each gender survey year the weighted number of individuals also needs to be known Note it is standard NTS methodology to include all respondents in the denominator not just those making a trip Select the Individual sav file Before producing the tabulation of gender by survey year the weighting must be specified From the menu select Data gt Weight cases Check Weight cases by and move across W2 household non response weight as the frequency variable Select OK The tabulation can now be specified From the menu select Analyze gt Descriptive Statistics gt Crosstabs Move Survey Year across to the row specification box Move Sex_B01ID across to the column specification box Select Cells and in the Non integer weights section select No adjustments Select OK The following SPSS output should be produced Survey year act
27. rip JTTXSC 1 TripTravTime JD JD Sum SD Stage distances NTS data extract 2002 2012 user guide Page 10 of 32 Non response and drop off weighting In addition to the short walk weights described above a weighting strategy for the NTS was developed following a recommendation in the 2000 National Statistics Quality Review of the NTS For the first time the 2005 NTS results were based on weighted data The weighting methodology was applied to data back to 1995 and all NTS figures for 1995 onwards which are published or released are now based on weighted data As well as adjusting for non response bias the weighting strategy for the NTS also adjusts for the drop off in the number of trips recorded by respondents during the course of the travel week for uneven recording of short walks by day of the week and for the short fall in reporting of long distance trips Further information on the weighting methodology is available in the 2012 Technical Report at the link below Please note that W notation in the Technical Report is not the same as the W notation in the NTS datasets as described in the Weighting matrices section below https www gov uk government uploads system uploads attachment_data file 225735 nts201 2 technical pdf Therefore there are several sets of weights which apply to different levels of the database household trip and long distance journey It is important t
28. rmine the unweighted sample size for trip rate analysis apply w1 to the trip data and w1 to the individual data Diary sample e To determine the unweighted sample size for household car ownership or driving licence figures apply no weights or wO Interview sample Long distance journey LDJ weights Long distance journeys trips over 50 miles are collected in two ways in the NTS via the travel diary and in the interview From 2002 2005 long distance trips were collected during the interview for an additional period of 3 weeks prior to the start of the travel diary From 2006 onwards they have been collected during the single week prior to the placement interview Analysis of the LDJ data a few years ago highlighted concerns over the quality of the interview recall data compared to the data collected during the travel week via the diary Therefore the LDJ weighting methodology was revised back to 2006 to take account of the a ee SSS SS SS SS eee NTS data extract 2002 2012 user guide Page 13 of 32 systematic under reporting of shorter long distance journeys The re weighting had little impact upon the overall LDJ trip rates However it resulted in a downward adjustment to distance estimates Revisions The following are the main changes made to the 2002 2012 datasets which account for the difference in the datasets for 2002 2010 previously available from the UK Data Service SQL processing of 2008 2012 raw data Due to cha
29. short walks and excluding series of calls trips JTTXSC Trip travelling time grossed for short walks and excluding series of calls These weighted variables have been constructed as follows SSXSC If series of calls SSXSC 0 If not series of calls and short walk stage SSXSC 7 If not series of calls and not short walk SSXSC 1 stage T A NTS data extract 2002 2012 user guide Page 9 of 32 STTXSC If series of calls STTXSC 0 StageTime If not series of calls and short walk stage STTXSC 7 StageTime If not series of calls and not short walk stage STTXSC 1 StageTime SD If short walk stage SD 7 StageDistance If not short walk stage SD 1 StageDistance JJXSC If series of calls JJXSC 0 If not series of calls and short walk trip JJXSC 7 If not series of calls and not short walk trip JJXSC 1 JOTXSC If series of calls JOTXSC 0 TripTotalTime If not series of calls and short walk trip JOTXSC 7 TripTotalTime If not series of calls and not short walk trip JOTXSC 1 TripTotalTime JTTXSC If series of calls JTTXSC 0 TripTravTime If not series of calls and short walk trip JTTXSC 7 TripTravTime If not series of calls and not short walk t
30. t Valid Percent Percent 8 0 8 0 8 0 16886 19467 19199 19904 19490 19735 18983 19914 19072 18069 19154 209873 In addition the unweighted number of stages for each mode should be obtained to ensure that the sample sizes on which the miles are based are sufficient Annex C Example tabulation 3 Multi coded data Reasons for not learning to drive by age Load the individual level data file Individual sav Before producing the tabulation of individuals and their reasons for not learning to drive by age the weighting must be specified E SS ES SSS NTS data extract 2002 2012 user guide Page 28 of 32 From the menu select Data gt Weight cases Check Weight cases by and move across W3 interview sample weight as the frequency variable Select OK In this case we are only interested in data for 2012 To restrict the data file to cases for 2012 From the menu select Data gt Select cases Check If condition is satisfied and select If Move SurveyYear across to the definition box and enter 2012 The box should read SurveyYear 2012 Select OK only Any outputs will now only use individuals from the 2012 survey The tabulation can now be specified From the menu select Analyze gt Descriptive Statistics gt Crosstabs Move ResNDN_A_BO1ID across to the row specification box Repeat this step for all ResNDN variables B to P Move Age_BO4ID
31. there is a weighting strategy to adjust for non response and drop off appropriate weights should be applied to all analyses of data See section on Non response and drop off weighting e All analyses of data from the travel diary should be based on fully co operating households only Analysis at the individual household and vehicle level should be based on the interview sample which includes data from additional partially co operating households See section on Weighting e Before carrying out any analysis all users should first try to reproduce the examples in the annexes of this document and some published results tables to ensure they are using the data and applying the weights correctly The 2012 NTS publication is available at https www gov uk government publications national travel survey 2012 e In addition users should be familiar with the survey methodology to ensure data is analysed and interpreted correctly See details below and the NTS technical information available at https www gov uk government collections national travel survey statistics e When analysing the data users should be aware that although the core survey has remained consistent over time some relatively minor changes are made to the survey each year e g addition removal of questions changes to filters and variable categories The appendix of the annual Technical Reports see link above give details of questionnaire changes s
32. ual year Sex of person Crosstabulation Count Sexofperson lt of Sexofperson lt Survey year actual year 8549 647 8944 077 17493 723 9578 418 9999 596 19578 014 9451 343 9850 284 19301 626 9843 876 10259 522 20103 397 9699 779 10093 774 19793 553 9772 677 10167 305 19939 982 9454 678 9798 554 19253 233 9770 212 10147 339 19917 551 9435 616 9774 031 19209 647 9008 588 9330 581 18339 169 9604 327 9942 373 19546 700 104169 160 108307 436 212476 596 EE NTS data extract 2002 2012 user guide Page 20 of 32 The trip counts can now be divided by the individual sample It should be noted that the trip counts are for a single week so therefore need to be multiplied by 52 14 number of weeks in a year 170 087 8 549 647 e g Trips per male in 2002 x 52 14 1 037 The following results should be calculated showing trips per person per year by gender Year Trip rate Male 2002 1037 2003 1027 2004 1013 2005 1031 2006 1014 2007 950 2008 969 2009 947 2010 938 2011 935 2012 935 2002 12 average 981 Female 2002 1055 2003 1040 2004 1039 2005 1056 2006 1060 2007 994 2008 1014 2009 997 2010 988 2011 981 2012 972 2002 12 average 1018 Total 2002 1047 2003 1034 2004 1026 2005 1044 2006 1037 2007 972 2008 992 2009 972 2010 963 2011 958 2012 954 2002 12 average 1000 C NTS data extract 2002 2012 user guide Page 21 of 32 Finally repeat the above step usi
Download Pdf Manuals
Related Search