ArticlePDF Available

Changes in the Representativeness of Strava Bicycling Data during COVID-19


Abstract and Figures

COVID-19 prompted a large global increase in Strava app use that could have implications for the representativeness of Strava bicycling data. We evaluated change in Strava representation of the general bicycling population from 2019 to 2020 by correlating Strava to counts of all bicyclists in Vancouver and Victoria, Canada. Strava data became more representative during COVID-19 lockdowns, likely because more bicycling activities were logged on the app. Age and gender bias in Strava data decreased and demographics of app users became more representative of the general population. For cities looking to increase Strava data reliability we suggest a benchmark Strava usage rate of ~7% of the total bicycle volumes.
Content may be subject to copyright.
Changes in the Representativeness of Strava Bicycling Data during
Jaimy Fischer
1 a , Trisalyn Nelson
2 , Meghan Winters
1 Faculty of Health Sciences, Simon Fraser University, 2 Department of Geography, University of California Santa Barbara
Keywords: active transportation, bicycling, big data, crowdsourced, COVID-19, Strava
UAI*,(&*.%,%(%!',-!'.,0**/-. .(/% 0
!&*%!.!('-(,. ,*,-'..!0'--(.,0!3%!'.L0%/.
 '!'.,0,*,-'..!('(. ',%!3%!'*(*/%.!(',(&B@AI.(
%!$%3/-&(,!3%!'.!0!.!-1,%(('. **L'',
!-!'.,0.,-'&(,* !-(**/-,-&&(,
,*,-'..!0(. ',%*(*/%.!('L(,!.!-%(($!'.(!',-.,0.
,%!!%!.31-/-.' &,$.,0/-,.(^G_(. .(.%!3%
1. Questions
.,0!--(!%8.'--**. .*(*%/-.(.,$. !,!3%!'.!0!.!-L
.,0.K',.,(&,(1-(/,.!0!.3.,$!'('. .,08.'--
**K*,(0!!3%,!,- !*. 1!.  /'*,'.-*.!%'.&*(,%
,-(%/.!('' . . , /-!'  ,(1!''/&, ( !3%!',-, K
*%''!'K' 0(3*,(#.-Q '', B@BAN %-('K ,-.,K. %L
B@BARL.,0. 0- (1'&(,..(-.,('(,,%.!('-1!. . ',%
!3%!'*(*/%.!(' Q ' ', B@BARL !- ,%. .( 1 ( /-- .,0
.(.,$ . !, ,!- !- 1%%(/&'. Q ' ', B@BARK1!.  .  .
.'!' .( /',,*,-'. 1(&' ' (%, /%.- Q,!6' ' !( B@AIRL
-, ('-(!(&(,* !,*,-'..!('(/.-!('',QLLK
(%%.-/  ,.,!-.!-K/.. %!&!.1(,$0!%%-/-.-. .8.'--
**/-,-,*,(&!''.%3 !.'&(,7/'.Q,,K.$!'-K'
,&,B@AIRS.,' . .&!,,(,- . -(!(&(,* ! ,.,!-.!- (
1 (!$-!''Q!,. .%LB@BARL.K. ,(/ . *'&!.!&&'3
&(, *(*% 0-.,. /-!' .,0K 1!.  ** /-,- !* ,!-!' CC_ .( GC
B@AI.(B@B@ Q.,0B@B@RL - /!',--*,(0($/,. ,+/-.!('-
(/. '-!'.,*,-'..!0'--L/,(%!-AR.(0%/.. !&*.-
Corresponding author: Corresponding author:
Jaimy Fischer
Fischer, Jaimy, Trisalyn Nelson, and Meghan Winters. 2022. “Changes in the
Representativeness of Strava Bicycling Data during COVID-19.”
, March.
(!',-.,0**/-('!.3U%0%(,,%.!('1!. (/'.-(%%!3%!'
'BR.(!'.!3 '-!'(,,%.!('!'-/-.-(.,0.,*,-'.!'
.,!**/,*(-'&(,* !-L
2. Methods
Study area
  -./3 , !- !' .  1-. (-. ( ' ' !'%/- .  !.3 (
'(/0, ' ,., !.(,! Q!'%/!' !.(,!K$ 3K '! K '
-+/!&%. &/'!!*%!.!-RK K ' Q..!-.!- ' B@AFRL   (-
. - !.!- /- (.   (6!% !$ (/'.-K &-/,- ( %% !3%
,!,- !*. .,'-' &,$.(---- '!',*,-'..!0'--
(!3%!'L(. !.!-%-( 0-!&!%,%3 ! !3%!'.(1(,$&(- ,-
,%.!0.(*,(0!'!%''.!('%0,-' 0-/-.'.!%!.3!'0-.&'.
!'%%-'!%!.!-!3%!'!',-.,/./,/.!5,!'-!4'. 2.'.
Bicycle count data
,.!3%(/'..-(,. *,!(*,!%AU.(,CAQB@AI
' B@B@RL !.3 -.5 *,(0! (/'.- ,(& AI *,&''. !3% (/'.,
%(.!('- !' '(/0,K ' BI .&*(,,3 ' *,&''. %(.!('- !'
!.(,!L!.3(/'..1,*,(0!. (/,%3,.!('-L/-. 
.,0- (, .( -- !.3 . *,(8%- (' .  .(.% '/&, ( *(*%
'*/%!!3%!'.!0!.!-,(,('.,0'.((1'%(. .,0
!3% (/'. . /- !' '%3-!- Q.,0 .,( B@BBRL  - . 1,
*,(0!-,. (/,%3U%0%(/'.-(!3%!'.!0!.!-,(--%%%!'$-
!' .  -.,. '.1(,$L  &.  (/'.- ,(& (.  -(/,- 3 %(.!('K
.K' (/,(. 3K,&(0!'(-,0.!('-1!. &!--!'.L '1
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 2
%AL./3, ,.,!-.!-L
Location Location PopulationPopulationa a Temperature and precipitationTemperature and precipitationb b Bicycle facilitiesBicycle facilitiesc c Bicycle mode shareBicycle mode shareaa (%) (%)
People People Density (ppl/Density (ppl/
kmkm22) )
Winter avg. Winter avg.
temp (temp (°C)
Summer avg. Summer avg.
temp (temp (°C)
Annual precipitation Annual precipitation
(mm) (mm)
Total distance Total distance
(km) (km)
Density (km/Density (km/
1000 ppl) 1000 ppl)
City City B.C. B.C. Canada Canada
City of
631,486 5,492.0 4.1 18.0 1189.0 495.6 0.8 6.1 2.5 1.4
235,689 1,673.2 4.6 16.9 882.9 205.4 0.9 8.7 2.5 1.4
(*/%.!(''&(- ,.1,(.!',(&. B@AF'!''-/-Q..!-.!-'B@AFR
%!&..1,(.!',(&'0!,('&'.'Q'0!,('&'.'%!&. ''B@BAR
!3%!',-.,/./,1-(.!',(&(*'.(,'(/0,Q!.3('(/0,B@BBR'-.5,(&. *!.%!('%!-.,!.(,,.,!.(,!Q!'%/!'!.(,!K$3K'! K'-+/!&%.RL
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 3
%BL '!'.,0/-,.-'!.3U%0%(,,%.!('L
City City Year Year Strava app Strava app
use use
Trips logged to Trips logged to
Strava Strava
City-level Strava usage City-level Strava usage
rate rate
, city-level (pooled , city-level (pooled
sample) sample)
Vancouver 2019 Reference Reference 5.7% 0.42
2020 +48 % +41 % 12.1% 0.73
Victoria 2019 Reference Reference 7.0% 0.85
2020 +26 % +27 % 15.8% 0.87
((&*,.,'-.1'!.!-K1%/%.. *,(*(,.!('% '!'. 
,(& .  .,0 - (, Q.,0 .,( B@BBRL ( 0%/. (1 .,0
,!,- !*,*,-'. (/'.-( %% !3%!-.-K1&-/,.,0/- ,.-
'----(,,%.!('.1'.,0'(/'.-(%%!3%!-.-L .,0
/- ,. !- .  *,(*(,.!(' ( %% .,!*- (/'. . !.3 !$ (/'.,- . .
1,*./,!' .  .,0. Q& (U(,,,(- .%L B@BARK ' . 
(,,%.!('!'!.-1 . ,.,0!3%!-.-*,(0! ((**,(2!&.!('
( %% !3%!-.- Q ' ', B@BARL /-.- ( .,0 . &3  &(,
,*,-'..!0 ( .  ',% *(*/%.!(' ' ,!0 .,'- (-,0 !' . 
*((% -&*% Q%% (-,0.!('- (&!'R Q!- ,K %-('K ' !'.,-
&-/,-(,.  *((%-&*% (.,0,!,- !* ' (,-/-.-( .,0
.3.,!**/,*(-'&(,* !-L*!8%%3K1/-%%-!'. .,0
..(2.,..-/-.-(,,,.!(''(&&/.,!,- !*K'(,. 
-&*%(,!,- !*',.31(&''3(%,/%.-QEEZRL
3. Findings
 '/&,(*(*%/-!'.,0.(.,$ !3%!'.!0!.!- '. .(.%
0(%/&(!3%!'.!0!.!-,(,('.,0!',-!'(. !.!-K1!. 
. ,.-.!',--!''(/0,Q%BRL''(/0,**/-,- !*',%3
(/% QDH_ !',-R ' .,0 !3%!' .!0!.!- !',- 3 DA_K
1 ,- !' !.(,! ** /- ' .!0!.!- !',- 3 BF ' BG_K
!3%!-.-**,.(%!'$K1!. /-,.-^GUA@_*(--!%3(,,-*('!'
.(-.,('!.3U%0%(,,%.!('QR\[@LGRL(!%%/-.,.K. .,0/-,.
,(- ,(& ELG .( ABLA_ !' '(/0, ' !.3U%0% (,,%.!(' !',-
-/-.'.!%%3K #/&*!' ,(& @LDB !' B@AI .( @LGC !' B@B@L ,U
(,,%.!('1-%,3-.,('!'!.(,!1 '. .,0/-,.1-G_
QB@AIR'!',-&,!'%%3!'B@B@Q@LHE.(@LHGR0'-. .,0/-
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 4
!/,AL '!'(,,%.!('QB@AIUB@B@R.1'.,0'(/'.-(%%!3%!-.-L - %$%!'&,$-. 
0%/(,-.,('(,,%.!('Q@LG(, ! ,RL,U('U3,(,,%.!('!',-!'(. !.!-K/.'(/0, . 
,.-.!'-K1!. (,,%.!('#/&*!',(&@LDB!'B@AI.(@LGC!'B@B@L,U(,,%.!('1-%,3-.,('!'
!.(,!!'B@AI'!',-&,!'%%3!'B@B@L -/-.(.,0.',.31(&'&('-.,.-.,('
(,,%.!('-!'(. !.!-!'B@B@K'(,,%.!('!'. -/-.',.3(%,/%.-!',--/-.'.!%%3!'
!5,'-!'(,,%.!('3.,!**/,*(-'&(,* !-!&*%3. .-/-.-
( .,0 ,!,- !* &3  &(, ,*,-'..!0 ( .  ',% !3%!'
*(*/%.!('L''(/0,K%%.,0.-/-.-- (13,U('U3,!',-
!'(,,%.!('1!. (/'.-(%%!3%!'K'. -.,('-.(,,%.!('(0,%%
1-(,. -&*%(.,0,!,- !*',.31(&'Q@LGIRL'!.(,!
. -.,('-.(,,%.!('(/,,!'. -&*%(,!,- !*',.3(%,
/%.- Q@LHHRL ' (.  !.!-K %, !',-- !' (,,%.!(' QBHUAD@_R !' . 
,,.!('% ,!,- !* -&*% -/-.'.!. ,-,  &('-.,.!' . .
!3%!'(,,,.!('!',-,&.!%%31 '!.!-1'.!'.(%($(1'
/,!'. 8,-.10(. *'&!Q/ %,'/ ,B@BARL
!.  ! , !.3U%0%.,0 /- ,.-K 1%-( - -.,(', (,,%.!('-!'
.  . -/-.- (, 1(&' ' (%, /%.-K -/-.!' . - ! , /-
,.- &3  !'!.!0 ( .., ,*,-'..!(' ( 1(&' ' (%, /%.-
!'.  .L !-!' .,0. ,%..( 1 ( !- /-!'.  ** !&*.-.
,%!!%!.3Q'',B@BARL/,8'!'-!&*%3. .'',!-!'
.,0**/-&3 0,-!'. -!.!-'. ..,0.&
&(,,*,-'..!01 '**/--/,!'. 8,-.&('. -(. *'&!L
 -8'!'-%-(,9. (1!3%!'K-*!%%3(,,,.!('K!',--
(&K'*(%!3. .-/**(,.-.!0.,'-*(,..!(',%!$%3.(('.!'/Q0'
' !.%(2B@BARK /.. %('U.,& 5.-(UAI('!3%!'
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 5
'.,0/-,-.!%%/'$'(1'L2*..%-.-(&(. &(,* !
- !.- !' .,0 . .( *,-!-.K /. /./, -./!- 1(/% %* %,!3 . -
 ,(/  . !- '%3-!- 1 - (1 . . .,0 ,*,-'..!0'-- ( .  ',%
!3%!'*(*/%.!('!-3'&!'!'9/'3%0%(**/-K1 !  !-
'&(,,'.%3K2.,&1. ,0'.-Q.,0B@B@KB@BARRL!. ..,
,*,-'..!(' ( 1(&' ' (%, /%.- !' .,0 .K *%'',- '
,-, ,- &! . /- . - -/-.- &(, */,*(-/%%3 .( 0' .!0
.,'-*(,..!(' +/!.3 (%- - . - , *,!(,!.3 *(*/%.!('- (, !',-!'
!3%!'!''L-1%%K1 ''(/ 1(&''(%,/%.-,/-!'
.,0. 3&3-/-&*%(.,0.. .-.,*,-'.*(*/%.!('U
%0%!3%!'*..,'-L'(. !.!-K!',--!'.,0**/-(,,-*('
1!.  .., ,*,-'..!(' Q!LLK ! , (,,%.!('R ( .  ',% !3%!'
*(*/%.!('K /. /,. , !'- &3  &,!'% 1 ' .  *,(*(,.!(' ( %%
!3%!'.,!*- %(3.,0, - -(&. ,- (%L(,!.!- %(($!'.(
!',-.,0 . ,*,-'..!0'--1-/-.  ' &,$ .,0 /-
,.(]^G_L !-,(&&'.!('(/%(%-.,3!.!('%-./!-
/-,.-K&*!'-. .'(/,!3%!-.-.(&*. !,,!-('.,0'
  /. (,- 1(/% %!$ .( $'(1% .,0 (, *,(0!!' .  . '
.!&%3,0!1( (/,1(,$' -.5..  *!.%!('% !-.,!.' . 
!.3('(/0,(,*,(0!!'/-1!. (6!%!3%(/'..L
/&!..J,/,3A@KB@BBK*.J, @IKB@BB
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0
International License (CCBY-SA-4.0). View this license’s legal deed at
licenses/by-sa/4.0 and legal code at for more
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 6
(--K,,'K,!-%3'%-('K '!'.,-K'(%!'L,-.,LB@AHLV-!',(1-(/,
..(('!.(, '!'*.!%..,'-(!3%!,- !*LWJournal of Transport & HealthI
/ %,K%* K'( '/ ,LB@BALVUAI&*.-('3%!'KB@AITB@B@LWTransport
& (U(,,,(-K,'!-(0!,K0!%(*!-U-.%%)K,!-%)*4U%('(K'
%,(,"LB@BALV'2&!'.!('(. .,0-.S,&.,.(-.!&.0,
''/%!%3!3%(%/&-('/,%(13-LWSafetyGQARJHL ..*-JOO(!L(,OA@LCCI@O
!.3('(/0,LB@BBLV!$13-LW ..*-JOO(*'.L0'(/0,LO2*%(,O.-.O!$13-O
(/./,K--LB@BALV9.!('-,(&. X.,0U* ,YJ/(-K(&&/'!.3K'Q%U
R/,0!%%'('(!%.1(,$(,. %.-LWQualitative Research in Sport, Exercise and Health
'0!,('&'.'%!&. ''LB@BALV'!'%!&.(,&%-LW(0,'&'.(
'L ..*-JOO%!&.L1. ,LLO%!&.P'(,&%-O!'2PL .&%L
!,. K!-%!'LK! %,'!('U%%-K '!'.,-K'L'',,!-LB@BALV (!$-M
'----&'.(!-/,'(&&/.!'!3%!',(&. '!'(&&/'!.3%. 
/,03LWFindingsK3L ..*-JOO(!L(,OA@LCBHFFO@@ALBBAFCL
!- ,K!&3K,!-%3'%-('K' '!'.,-LB@B@LV(&*,!'*.!%--(!.!('-(
(&&/.!'0,-/-,.!('%!,- !**./,3. .,0!.'--**LWFindingsGG@
Q*.&,RJATIL ..*-JOO(!L(,OA@LCBHFFO@@ALAFGA@L
,,K! %LK,!L.$!'-K'! %L,&,LB@AILV(&*,!'!3%!-.- (-
&,.* ('**-.((,!-1!.  (- (((.J&*%!.!('-(,*,-'..!0'--
'%.!('!-LWJournal of Transport & HealthAEQ&,RJA@@FFAL ..*-JOO(!L(,O
,!6'K,K'/''!(LB@AILV,(1-(/,!'!3%(%/&-J2*%(,!'. (%(
(%/'.,(,* !'(,&.!(''-.%!- ('!.(,!'. (-LWSocArXivK'/,3L
..*-JOO(!L(,OA@LCABCEO(-L!(OC L
K3/ 3/'K'*$-',LB@BALV.,0.,(.(,!3%('!.(,!'J!.,./,
%-('K,!-%3'K(%!',-.,K,',K'!%/%%,K' '!'.,-LB@BAL
V,(1-(/,.(,!3%!'-, ',.!LWTransport ReviewsDAQARJIGTAADL
%-('K,!-%3'K0!*-(3K(%!',-.,K!&3!- ,K'--,/&U-.(-K,',K
' '/K' '!'.,-LB@BALV',%!4(%(,**!'!3%!,- !*1!. 
,(1-(/,.LWTransportation Research Part C: Emerging TechnologiesABEQ*,!%RJA@BIHAL
..!-.!-'LB@AFLV'-/-,(8%KB@AF'-/-LW ..*-JOO111ABL-..'LLO'-/-U
.,0LB@B@LV,!'*(,.LW ..*-JOOA'D,'HH$D4! .GAC%E/U1*'!'L'.'U--%L(&O1*U
SSSLB@BALV,!'*(,.LW ..*-JOOA'D,'HH$D4! .GAC%E/U1*'!'L'.'U--%L(&O1*U
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 7
.,0.,(LB@BBLV(&LWB@BBL ..*-JOO&.,(L-.,0L(&OL
,.!!*.!('',0% 0!(/,J/%.!*,-*.!0!1LWJournal of Transport GeographyIE
 '-!'. *,-'..!0'--(.,0!3%!'./,!'UAI
Findings 8
... It is widely use in the runner's community who openly shares GPS recorded workouts. Strava popularity increased particularly during the COVID-19 pandemic as it offers, especially during lockdown periods, opportunities for people to be followed up, to participate in virtual challenges or competitions and to be connected to friends and peers' trough virtual community while social distancing and restrictions prevent group training and events (Couture, 2021;Fischer et al., 2022). One of the most innovating features of Strava is the concept of segments, which are portions of road or trail created by members where athletes can compare times. ...
Full-text available
This study aimed to assess the reliability of Strava measurements when manipulating segment distance and running velocity. The tests were carried out on a flat and straight segment. Ten male regular runners were equipped with a Garmin® Forerunner 945 watch and ran over a distance of 1 km of four increasing speeds: 1.39, 2.78, 4.17 and 5 m/s. Different reference positions were accurately determined in order to calculate time at 100 m, 200 m, 500 m, 700 m, and 1000 m. A bike with a wide angle camera was used to standardize the run pace and to record the entire run for reference measurements. Results show a high level of reliability with nearly perfect intra-class correlation (from .997 to 1) when data is analysed accordingly to the distance of the segment or to the running velocity. The validity is also very good with a small average bias (-0.25 s), a standard deviation of differences of 1.84 sec and the limit of agreement range from-3.86 to 3.35 sec. Regardless of the length of the segment, the actual performance of the runner is normally within +/-2 seconds of the results given by the Strava application. In 95% of cases, the measurement error will be less than four seconds. The relative error is proportionally larger for short segments done at a fast pace. Further studies are needed to explore Strava segments reliability in other specific contexts.
... Strava is utilized by those with access to the technology, which excludes people without smart devices. Strava users are often younger men who tend to have higher SES, and individuals with low SES are likely underrepresented [47,79,85]. Reaching underrepresented populations would allow for a more accurate account of the routes selected by runners from all backgrounds. ...
Full-text available
Running can improve physical health and psychological wellbeing. However, the characteristics of conducive running environments are relatively unknown. This study determines neighborhood factors that attract running and explores how age and gender mediate built environment preferences. Spatial patterns of runners in Metro Vancouver were identified using crowdsourced fitness data from Strava, a popular application for tracking physical activities. The influence of socio-economic status (SES), green and/or blue space, and urbanicity on route popularity was assessed using a Generalized Linear Model (GLM). The influence of these neighborhood variables was also calculated for runners by age and gender. The results show high neighborhood SES, the presence of green and/or blue space, and high population density are associated with increased running activities in all age and gender groups. This study contributes a novel approach to understanding conducive running environments by demonstrating the utility of crowdsourced data in combination with data about urban environments. The patterns of this large group of runners can be used to inform planning for cities that promote running, as well as seek to encourage equal participation among different ages and genders.
... Similarly, in the Netherlands a diligent advertising campaign for the Bike PRINT app was able to reached representative samples (Garber, Watkins, and Kramer 2019). During the COVID-19 lockdowns, the bias in Strava representativeness (age-wise and gender-wise) decreased in Vancouver and Victoria, Canada (Fischer, Nelson, and Winters 2022). This is might be the case in other locations as well as the pandemic encouraged more AT as a non-pharmaceutical intervention. ...
In the era of ubiquitous technology, crowdsourced data is an emerging frontier for active travel (AT) studies. In this work, we utilize accrued knowledge from interviews and previous literature regarding crowdsourced data strengths, challenges, usefulness and reliability for future informants who seek to embrace crowdsourced data. We review four main types of crowdsourced data: social fitness networks, in-house developed apps, bike sharing systems and participatory mapping. The strengths of crowdsourced data include providing fine data coverage, precision, details, immediacy and empowering users to participate in decision-making. Potential challenges that might arise from adopting this data are related to technical, privacy, proprietorship, financial and data fragmentation factors. In terms of usefulness, crowdsourced data lend themselves to before and after analysis, assessing current infrastructure, and investment prioritization. Reliability issues that may undermine the credibility of crowdsourced data are also discussed, as well as remedies for these concerns.
Full-text available
This paper discusses possible long-term effects of COVID-19 on activity-travel behaviour. Making use of theories and concepts from economics, psychology, sociology, and geography, this work argues that lasting effects can be expected, and specifically that peak demand among car and public transport users may be lower than if the pandemic would never have happened. The magnitude of such effects at the aggregate level in terms of the total travel time of all inhabitants of a country or state is likely limited. Such lasting effects imply that additional infrastructure extensions to reduce congestion on roads and crowding in public transport might have a lower benefit-cost ratio than would be the case without these impacts. The paper discusses avenues for future research, including work on the role of attitude changes, the formation of new habitual behaviour, new social norms and practices, well-being effects, and the role of Information and Communication Technologies (ICT).
Full-text available
We used 2013/2014 Canadian Community Health Survey to describe who bicycles for leisure, commuting, or both leisure and commuting. Nearly one-quarter of Canadians bicycled in the 3 months prior to the survey: 7 Canadians bicycled for leisure for every 1 person who bicycled for commuting purposes. People bicycling for leisure were more likely to be younger, male, higher income, and identify as white. Commute bicycling captured a very small proportion of the bicycling population; men were nearly twice as likely to commute compared to women and there was little difference in bike commuting across racial identity.
Full-text available
Fitness apps, such as Strava, are a growing source of data for mapping bicycling ridership, due to large samples and high resolution. To overcome bias introduced by data generated from only fitness app users, researchers build statistical models that predict total bicycling by integrating Strava data with official counts and geographic data. However, studies conducted on single cities provide limited insight on best practices for modeling bicycling with Strava as generalizability is difficult to assess. Our goal is to develop a generalized approach to modeling bicycling ridership using Strava data. In doing so we enable detailed mapping that is more inclusive of all bicyclists and will support more equitable decision-making across cities. We used Strava data, official counts, and geographic data to model Average Annual Daily Bicycling (AADB) in five cities: Boulder, Ottawa, Phoenix, San Francisco, and Victoria. Using a machine learning approach, LASSO, we identify variables important for predicting ridership in all cities, and independently in each city. Using the LASSO-selected variables as predictors in Poisson regression, we built generalized and city-specific models and compared accuracy. Our results indicate generalized prediction of bicycling ridership on a road segment in concert with Strava data should include the following variables: number of Strava riders, percentage of Strava trips categorized as commuting, bicycling safety, and income. Inclusion of city-specific variables increased model performance, as the R 2 for generalized and city-specific models ranged from 0.08-0.80 and 0.68-0.92, respectively. However, model accuracy was influenced most by the official count data used for model training. For best results, official count data should capture diverse street conditions , including low ridership areas. Counts collected continuously over a long time period, rather than at peak periods, may also improve modeling. Modeling bicycling from Strava and geographic data enables mapping of bicycling ridership that is more inclusive of all bicyclists and better able to support decision-making.
Full-text available
In Spain, a new challenge is emerging due to the increase of many recreational bicyclists on two-lane rural roads. These facilities have been mainly designed for motorized vehicles, so the coexistence of cyclists and drivers produces an impact, in terms of road safety and operation. In order to analyze the occurrence of crashes and enhance safety for bicycling, it is crucial to know the cycling volume. Standard procedures recommend using data from permanent stations and temporary short counts, but bicycle volumes are rarely monitored in rural roads. However, bicyclists tend to track their leisure and exercise activities with fitness apps that use GPS. In this context, this research aims at analyzing the daily and seasonal variability of the Strava Usage Rate (SUR), defined as the proportion of bicyclists using the Strava app along a certain segment on rural highways, to estimate the Annual Average Daily Bicycle (AADB) volume on rural roads. The findings of this study offer possible solutions to policy makers in terms of planning and design of the cycling network. Moreover, the use of crowdsourced data from the Strava app will potentially save costs to public agencies, since public data could replace costly counting campaigns.
Full-text available
In recent years, there has been a growing interest in the practice of digital self-tracking. Researchers have drawn attention to who self-tracks, why people self-track, and what it feels like to self-track in the context of sport and physical activity. To date, limited research has focused on self-tracking as a social practice and there has been minimal engagement with the specific online platforms that individuals use to share their self-tracking data online. In this paper I engage with findings from an ethnographic study of Strava, a popular social fitness platform. I propose that while Strava can be a source of motivation and entertainment for its users, and even help to establish or strengthen social networks, the platform invites users to adopt and adapt to technologically-mediated surveillance strategies that encourage and reward displays of bodily self-discipline.
Full-text available
Strava Metro data are used in bicycle planning, but there are concerns it overrepresents fitness activity. The data include a commute label, but spatial patterns of commuting versus recreational ridership are underexplored. Using spatial regression, we compare associations of Strava ridership by trip type. Commuting was associated with areas with more on-street infrastructure, universities, and higher bicycle crash density. Recreational ridership was higher in areas with older populations, more hills and major roads, and lower intersection density. Both trip purposes tended to be in areas with regional trails, off-street infrastructure, higher bicycle mode share, bridges, and proximity to the ocean.
Full-text available
Cities are promoting bicycling for transportation as an antidote to increased traffic congestion, obesity and related health issues, and air pollution. However, both research and practice have been stalled by lack of data on bicycling volumes, safety, infrastructure, and public attitudes. New technologies such as GPS-enabled smartphones, crowdsourcing tools, and social media are changing the potential sources for bicycling data. However, many of the developments are coming from data science and it can be difficult evaluate the strengths and limitations of crowdsourced data. In this narrative review we provide an overview and critique of crowdsourced data that are being used to fill gaps and advance bicycling behaviour and safety knowledge. We assess crowdsourced data used to map ridership (fitness, bike share, and GPS/accelerometer data), assess safety (web-map tools), map infrastructure (OpenStreetMap), and track attitudes (social media). For each category of data, we discuss the challenges and opportunities they offer for researchers and practitioners. Fitness app data can be used to model spatial variation in bicycling ridership volumes, and GPS/accelerometer data offer new potential to characterise route choice and origin-destination of bicycling trips; however, working with these data requires a high level of training in data science. New sources of safety and near miss data can be used to address underreporting and increase predictive capacity but require grassroots promotion and are often best used when combined with official reports. Crowdsourced bicycling infrastructure data can be timely and facilitate comparisons across multiple cities; however, such data must be assessed for consistency in route type labels. Using social media, it is possible to track reactions to bicycle policy and infrastructure changes, yet linking attitudes expressed on social media platforms with broader populations is a challenge. New data present opportunities for improving our understanding of bicycling and supporting decision making towards transportation options that are healthy and safe for all. However, there are challenges, such as who has data access and how data crowdsourced tools are funded, protection of individual privacy, representativeness of data and impact of biased data on equity in decision making, and stakeholder capacity to use data given the requirement for advanced data science skills. If cities are to benefit from these new data, methodological developments and tools and training for end-users will need to track with the momentum of crowdsourced data.
Full-text available
Increasing population levels of cycling has the potential to improve public health by increasing physical activity. As cyclists have begun using smartphone apps to record trips, researchers have used data generated from these apps to monitor cycling levels and evaluate cycling-related interventions. The goal of this research is to assess the extent to which app-using cyclists represent the broader cycling population to inform whether use of app-generated data in bike-infrastructure intervention research may bias effect estimates. Using an intercept survey, we asked 95 cyclists throughout Atlanta, Georgia, USA about their use of GPS-based smartphone apps to record bike rides. We asked respondents to draw their common bike routes, from which we assessed the proportion of ridership captured by app-generated data sources overall and on types of bicycle infrastructure. We measured socio-demographics and bike-riding habits, including cyclist type, ride frequency, and most common ride purpose. Cyclists who used smartphone apps to record their bike rides differed from those who did not across some but not all socio-demographic characteristics and differed in several bike-riding attributes. App users rode more frequently, self-classified as stronger riders, and rode proportionately more for leisure. Although groups had similar infrastructure preferences at the person level, differences appeared at the level of the estimated ride, where, for example, the proportion of ridership captured by an app on protected bike lanes was lower than the overall proportion of ridership captured. A sample calculation illustrated how such differences may induce selection bias in smartphone-data-based research on infrastructure and motor-vehicle-cyclist crash risk. We illustrate in the sample scenario how the bias can be corrected, assuming inverse-probability-of selection weights can be accurately specified. The presented bias-adjustment method may be useful for future bike-infrastructure research using smartphone-generated data.
Monitoring bicycle trips is no longer limited to traditional sources, such as travel surveys and counts. Strava, a popular fitness tracker, continuously collects human movement trajectories, and its commercial data service, Strava Metro, has enriched bicycle research opportunities over the last five years. Accrued knowledge from colleagues who have already utilised Strava Metro data can be valuable for those seeking expanded monitoring options. To convey such knowledge, this paper synthesises a data overview, extensive literature review on how the data have been applied to deal with drivers’ bicycle-related issues, and implications for future work. The review results indicate that Strava Metro data have the potential—although finite—to be used to identify various travel patterns, estimate travel demand, analyse route choice, control for exposure in crash models, and assess air pollution exposure. However, several challenges, such as the under-representativeness of the general population, bias towards and away from certain groups, and lack of demographic and trip details at the individual level, prevent researchers from depending entirely on the new data source. Cross-use with other sources and validation of reliability with official data could enhance the potentiality.