Table 1 Non-Supervised Physical Activity Programs (NSPAP) in the workplace

1. Marshall AL. Print vs website PA programs: a randomized trial. 2003

Behavioral program with dissemination of educational materials (strategies); motivational program through the dissemination of motivational messages; non-individualized content, no counseling session

Interventional, comparative, randomized study

Comparison among the use of paper-based (letters, booklet) vs computer-based messages (mails, recalled website)

N = 655 (office workers)

Allocated to Print intervention: N = 328

Allocated to Web intervention: N = 327

8 weeks

TTM

2. Proper KI. Effect of individual counseling on PA fitness and health: a RCT in a workplace setting. 2003

Individualized informative and behavioral program through counseling sessions (strategies, goals, information)

Interventional, comparative, randomized study

Comparison among the entire incentive program vs written information about lifestyle factors only

N = 299

(3 workplaces, municipal office workers) IG: incentive program N = 131

CG: information N = 168

9 months

TTM

3. Chan CB. Health benefits of a pedometer-based PA intervention in sedentary workers. 2004

Motivational program with the use of a pedometer; individualized behavioral program with the setting up of goals and strategies, counseling sessions; and informative program with educational materials during working groups

Interventional, non-comparative study

N = 177

(public administration office workers)

12 weeks

No theoretical framework

4. Hallam JS. The Long-Term Impact of a Four-Session

Work-Site Intervention on Selected SCT Variables Linked to Adult Adherence. 2004

Instructional intervention

composed of four 60-min sessions delivered across 2 weeks, to assess SCT variables linked to exercise behavior

Three SCT variables were measured for all observations: (1) outcome-expectancy

value, (2) exercise self-efficacy, and (3) the use of self-regulation strategies for exercise

N = 210

(service-type industry)

IG: N = 95

CG: N = 115

12 months

SCT

5. Plotnikoff RC. Efficacy of an E-mail intervention for the promotion of PA and nutrition behavior in the workplace context. 2005

Motivational program: dissemination of non-individualized motivational messages

Interventional, comparative, randomized study. The IG received one PA and one parallel nutrition message per week for 12 weeks. The CG received no weekly messages

N = 2 121

(5 workplaces)

IG: N = 1566

CG: N = 555

12 weeks

SCT, TTM,

Theory of Planned Behavior

6. Griffin-Blake CS. Evaluation of social-cognitive vs stage-matched, self-help PA interventions at the workplace. 2006

Individualized (depending on the group), informative, behavioral and motivational program through the dissemination of educational materials (awareness, strategy, strengthening of motivation)

Interventional, comparative, randomized study. Comparison among the program adjusted individually vs the generic program (no individual adjustment): participants were randomly assigned into either (a) the self-help exercise program based on the TTM or (b) the self-help exercise program based on SCT

N = 208

(university staff)

TTM group: N = 115

SCT group: N = 93

1 month

TTM vs SCT

7. Plotnikoff RC. The efficacy of stage-matched and standard public health materials for promoting PA in the workplace: the PA Workplace Study (PAWS). 2007

Individualized, informative, behavioral and motivational program through the dissemination of educational materials, awareness, strategy, strengthening of motivation (TTM group) vs standard group

Interventional, comparative, randomized study. Comparison among the TTM program adjusted individually (stage-matched group) vs the generic program (no individual adjustment). Five motivationally targeted booklets were developed for the stage-matched group. The standard group received PA Guide and handbook

N = 507 (3 companies, office workers)

Staged match group: N = 115

Standard intervention (CPAG): N = 176

CG: N = 166

12 months

TTM

8. Opdenacker J. Effectiveness of face-to-face vs telephone support in increasing PA and mental health among university employees. 2008

Individualized behavioral program, based on counseling sessions (goal and strategy setting up); informative program through the dissemination of educational materials (brochure)

Interventional, comparative, randomized study

Comparison of a behavioral program delivered by phone vs by face-to-face interview

N = 90 (university staff)

Allocated to face-to-face group: N = 33

Allocated to telephone support: N = 33

3 months

SCT

09. Dishman RK. Move to Improve: A Randomized Workplace Trial to Increase PA. 2009

Informative program with educational materials; behavioral program with setting up of goals and strategies; motivational program with pedometer, material rewards, team building with challenges

Interventional, comparative, randomized study

Comparison among the entire incentive program vs the informative program alone (newsletter)

N = 1 442

(16 workplaces, office workers)

IG: N = 885

CG: N = 557

12 weeks

Goal-setting theory

10. Dishman RK. Dose relations among goal setting, theory-based correlates of goal setting and increases in PA during a workplace trial. 2010

Nested study, ref. n°9 (Dishman RK, 2009)

Originality of the study: this study examined mediators/moderators of the outcomes of PA interventions and the dose relation of goal setting with PA

N = 1 442

(16 workplaces, office workers)

IG: N = 885

CG: N = 557

12 weeks

Goal-setting theory

11. Samuels TY. A randomized controlled trial of continuous activity, short bouts, and a 10,000 steps guideline in inactive adults. 2011

Behavioral program through setting up of daily PA goals; motivational program through the use of a pedometer; no individual adjustment; no counseling session

Interventional, comparative, randomized study

Comparison among groups with goal of 10,000 steps/day + pedometer vs 30 min of activity per day vs at least 3 sessions of 10 min of PA

N = 50

(university employees)

3 groups are compared after randomization:

G1: 10. 000 steps: N = 18

G2: 30 mn of PA: N = 17

G3: 3 × 10 mn of PA: N = 15

5 weeks

SCT

12. Mc Eachan RRC. Testing a workplace PA intervention: a cluster randomized controlled trial. 2011

Informative program with educational materials; motivational program through encouraging messages and setting up of challenges; implementation of the program by facilitators

Interventional, comparative, randomized study

Comparison among the incentive program vs the CG

N = 1 260

(5 public organizations,

44 worksites)

Intervention worksites: 22 G: N = 668

CG: N = 606

3 months

Theory of planned behavior

13. Robroek SJW. Cost-effectiveness of a long-term Internet-delivered

worksite health promotion programme on physical

activity and nutrition: a cluster RCT. 2012

Incitative program using the internet to promote health behavior change: extensive computer-tailored advice on self-reported PA and fruit and vegetable intake

Interventional, comparative, randomized study

Comparison among the incentive program vs the CG

N = 924

IG: N = 465

CG: N = 459

12 months

No theoretical framework

14. Gazmararian JA

A randomized prospective trial of a worksite intervention program to increase PA. 2013

Allowance of paid time for practicing PA; support for a fitness club membership

Interventional, comparative, randomized study. Comparison among different combinations of the incentive program (4 different groups) vs the CG

N = 410

60 university departments randomized into five groups, with a CG (N = 70)

9 months

SCT

15. Hunter RF. PA loyalty cards for behavior change: a quasi-experimental study. 2013

Motivational program using a pedometer in the workplace, financial rewards (PAL scheme); behavioral program with goal setting up and educational materials (website)

Comparison among the program with vs without rewards. In the CG, participants used a PA loyalty card (PAL card) to self-monitor PA levels over the 12-weeks intervention period but did not collect points or earn rewards

N = 406

(2 buildings, office workers)

Incentive group: N = 199

CG: N = 207

6 months

Learning theory

16. Van Hoecke AS. Long-term effectiveness and mediators of a need-supportive PA coaching among Flemish sedentary employees. 2013

Individualized behavioral program, based on counseling sessions, with setting up of goals and strategies

Interventional, comparative, randomized study

Comparison among the incentive program (sedentary employees guided by a personal PA coach) vs a CG

N = 122

(university employees)

Coaching group: N = 92

CG: N = 34

4 months + one-year follow up

Self-determination theory

17. Viester L. Process evaluation of a multifaceted health program aiming to improve PA levels and dietary patterns among construction workers. 2014

Individualized behavioral program through counseling sessions (strategies, goals); motivational program with use of a pedometer; informative program (dissemination of educational materials)

Interventional, comparative, randomized study

Comparison among the incentive program (face-to-face and phone coaching contacts provided by personal health coaches (PHCs)) vs a CG

N = 314

(construction workers)

IG: N = 162

CG: N = 152

6 months

TTM

18. Bale JM

Effect of the Work Environment on Using Time at Work to Exercise. 2015

Interventional RCT with quality control. Allowance of paid time for practicing PA on workplace

In two groups, every individual received both a gym membership and 30 min for exercise per workday. One of the two groups received education materials throughout the study period. Guidelines for taking the 30 min were explained in oral, written, and e-mail format

N = 188

30 university departments randomized on two groups

Gym + Time: N = 129

Gym + time + educ: N = 119

9 months

SCT

19. Lippke S

A Computerized Lifestyle Application to Promote Multiple Health Behaviors at the Workplace: Testing Its Behavioral and Psychological Effects. 2015

Individualized, informative, behavioral and motivational program through the dissemination of educational materials (awareness, strategy, strengthening of motivation)

Interventional, comparative, randomized study

Comparison among the entire individualized program vs informative program (generic educational materials)

N = 560

(shiftworkers, track workers, train drivers, ticket inpectors, 45% with a physical occupation)

Stage matched intervention: N = 498

Active control condition: N = 62

1 month

TTM

20. Macniven R. Does a corporate worksite PA program reach those who are inactive? Findings from an evaluation of the Global Corporate Challenge. 2015

Motivational program: use of a pedometer and achievement of a challenge (as a team), motivational messages

Interventional, non-comparative study

Assessment of the evolution of the PA level within the study population

N = 587 (office workers)

No CG

16 weeks

No theoretical framework

21. Mainsbridge CP. The Effect of an e-Health Intervention Designed to Reduce Prolonged Occupational Sitting on Mean Arterial Pressure. 2015

Educational and interactive

e-health software program, prompting employees to engage in a brief bout of short burst PA periodically (13 weeks)

RCT, e-health intervention

Primary outcome: Mean Arterial Pressure (MAP)

N = 29 (desk-based employees from a Department of Police and Emergency Management)

IG = 11

CG = 18

13 weeks

TTM

22. Mansi S. Investigating the effect of a 3-month workplace-based pedometer-driven walking programme on health-related quality of life in meat processing workers: a feasibility study within a RCT. 2015

Intervention participants utilized a pedometer and educational material, a step calendar to self-monitor their activity

Single-blinded RCT

Participants were required to walk to accumulate at least 30 mn of moderate intensity activity for at least 5 days/week during work and/or leisure time

N = 58 (large meat processing plant)

IG = 29

CG = 29

12 weeks

Self-Regulation Theory

23. Lawton R. Intervention fidelity and effectiveness of a UK worksite physical activity intervention funded by the BUPA Foundation, UK. 2015

Fidelity analysis as part of a large matched-pair cluster randomized controlled trial of a worksite physical activity intervention (AME for Activity)

The nine key components of the intervention called ‘AME (Awareness, Motivation and Environment) for activity’ were: a launch week, interactive leaflets, posters, a knowledge quiz, team challenges, reminders, letters of management support, newsletters and fridge magnets to allow self-monitoring of PA

N = 1260

IG = employees in the Local Council (N = 443)

CG = employees in 4 other worksites: hospital, bus company, government organization, university

(N = 611)

3 months + 9 months follow up

Theory of planned behavior

24. Finkelstein EA. Effectiveness of activity trackers with and without incentives to increase PA (TRIPPA): a RCT. 2016

Motivational program with the use of a pedometer and financial rewards (depending on the group); individualized informative and behavioral program through the dissemination of educational materials (awareness, strategies, goals)

RCT. Participants were assigned to one of four study groups: control (no tracker or incentives), activity tracker and website (Fitbit), tracker plus charitable incentives (charity), and tracker plus cash incentives (cash)

N = 800

(15 companies, office workers)

Fitbit: 203

Charity: 199

Cash incentives: 197

CG: 206

6 months + 6 months follow-up

No theoretical framework

25. Carr LJ. Total Worker Health Intervention Increases Activity of Sedentary Workers

2016

Ergonomic Program promoting PA, using a portable seated elliptical machine, an i-Pod Touch to track participant’s daily pedaling progress

Overweight/obese adults working in sedentary desk jobs were randomized

to: (1) a health protection–only group (HPO, n = 27); or (2) an integrated health protection/health promotion group (HP/HP, n = 27)

N = 60

(Large private Company)

IG (HP/HP) = 30

CG(HPO) = 30

16 weeks

SCT

26. Chaélat-Valayer E. Long-term effectiveness of an educational and physical intervention for preventing low-back pain recurrence: a RCT. 2016

RCT investigating the effect of a light exercise program, initiated in the workplace and continued at home, in reducing recurrence of LBP (Low Back Pain) episodes among healthcare workers

The intervention comprised three steps: (i) a 2-h education session about LBP (Low back Pain) prevention, (ii) five weekly 90-min exercise training sessions in the workplace, and (iii) a home-based self-managed exercise program

N = 342

(Health care workers from ten hospitals)

IG = 171

CG = 171

2 years

No theoretical framework

27. Taylor WC. Impact of Booster Breaks and Computer Prompts on PA and Sedentary Behavior Among Desk-Based Workers: A Cluster-RCT. 2016

RCT investigating the effect of a Booster Break program. The Booster Break was implemented with an exercise physiologist, PA specialist with behavioral sciences training, instructional designer, health educator, and videographer

3-armed, cluster RCT at 4 worksites. IG1 = Booster Break arm: stretching, strengthening, and aerobic movements, followed by a 60-s meditation. Daily worksite sessions lasted 13 to 15 min during one 15-min break

IG2 = computer-prompt arm

N = 185

(office workers)

IG1 = 76

IG2 = 61

CG = 48

6 months

SCT

28. Aittasalo M. Moving to business—changes in PA and sedentary behavior after multilevel intervention in small and medium-size workplaces. 2017

Information campaign set up by working groups (company employees)

Interventional, non-comparative study

Assessment of the evolution of the PA level of the participants

N = 396

12 companies in all sectors with less than 250 employees

1 year

No theoretical framework

29. Raedeke TD. High Vs Low Theoretical Fidelity Pedometer Intervention Using Social-Cognitive Theory on Steps and Self-Efficacy. 2017

Motivational program: use of a pedometer, with the formation of walking groups; behavioral program with goal and strategy setting up within the walking groups

RCT, IG (High Fidelity Program): use a pedometer, weekly group walk followed by a meeting to discuss cognitive behavioral strategies targeting self-efficacy. CG (Low theoretical Fidelity Program): met for a group walk, use a pedometer as a motivational tool and to monitor steps

N = 62

(university employees)

Allocated to High Fidelity Program (N = 32)

Allocated to Low Fidelity Program (N = 30)

10 weeks

SCT

30. Losina E. Implementation of a workplace intervention using financial rewards to promote adherence to physical activity guidelines: a feasibility study. 2017

Financial incentive program based on SCT theory: B. and Women’s Wellness (B-Well) program. No control arm

Secondary outcomes included Fitbit-wear adherence and factors associated with meeting CDC guidelines more consistently

Participants were rewarded for increasing their MVPA by 10% from the previous week or for meeting the Centers for Disease Control and Prevention (CDC) PA guidelines (150 min of MVPA per week). Primary outcome was the % of participants meeting weekly MVPA goals and CDC PA guidelines

N = 300

(sedentary hospital employees)

26 weeks

SCT

31. Hunter RF. Association among time preference, present-bias and PA: implications for designing behavior change interventions. 2018

Nested study. (Ref n°15; Hunter RF, 2013)

Specific objective was to determine whether the"present time"interest may influence the PA level

Comparative, randomized study. PAL scheme vs CG

“Discount rate” and “present-bias” parameters are studied

N = 176

(2 buildings, office workers)

IG (PAL-scheme): N = 95 from [N = 199]

New CG: N = 111

12 weeks

Learning theory

32. Hunter RF. Effectiveness and cost-effectiveness of a loyalty scheme for PA behaviour change maintenance: results from a cluster RCT. 2018

Nested study (Ref n°15; Hunter RF, 2013)

Specific objective was to determine cost-effectiveness of the PAL loyalty scheme

Interventional, comparative, randomized study

Comparison among the program with vs without rewards

N = 853

(9 workplaces, 27 clusters,

office workers)

IG: N = 457

CG: N = 396

12 months

Learning theory

33. Reed JL. The Impact of Web-Based Feedback on PA and Cardiovascular Health of Nurses Working in a Cardiovascular Setting: A Randomized Trial. 2018

Motivational program with the use of a pedometer and setting up of challenges

Participants were asked at baseline and every 2 weeks to self-set goals to increase daily pedometer steps and the weekly number of 10-min blocks of MVPA

Interventional, comparative, randomized study

Comparison of the program completion as an individual vs as a group of friends vs as a team of colleagues

N = 76

(care setting: nurses)

Allocated to Individual intervention: N = 25

Allocated to Friend intervention: N = 25

Allocated to Team intervention: N = 25

6 weeks

Self-determination continuum theory

34. Park J. Motivational Interviewing for Workers with Disabling Musculoskeletal Disorders: Results of a Cluster Randomized Control Trial

2018

Cluster RCT with

claimants attending an occupational rehabilitation facility

Six clinicians provided Motivational Interviewing (MI) in addition to the standard functional restoration program and formed an intervention group. Claimants were predominantly employed (72.7%), males (63.2%), with moderate levels of pain and disability (mean pain VAS = 5.0/10

and mean Pain Disability Index = 48/70)

N = 728

(injured workers receiving workers’ compensation and undergoing work rehabilitation)

IG: N = 367

CG: N = 361

7 months

Motivational Interviewing theory

35. Gremaud AL. Gamifying Accelerometer Use Increases Physical Activity Levels of Sedentary Office Workers. 2018

The purpose of this study was to test the efficacy of MapTrek for increasing daily steps and moderate-intensity steps over 10 weeks in a sample of sedentary office workers

MapTrek is a mobile health platform that gamifies Fitbit use for the purpose of promoting physical activity. Each participant received a Fitbit Zip to wear daily throughout the intervention. Participants were randomized to either a: Fitbit-only group (FB) or Fitbit + MapTrek group (MT)

N = 146

(office workers)

IG: N = 73

CG: N = 72

10 weeks

Self-determination continuum theory

36. Murray JM. Predicting Outcomes from Engagement With Specific Components of an Internet-Based PA Intervention With Financial Incentives: Process Analysis of a Cluster Randomized Controlled Trial. 2019

Nested study. Ref n°15, Hunter RF, 2013

Specific objectives were to determine: (1) whether engagement in specific

intervention components predicted PA, (2) targeted mediators, (3) predictors of website non usage

Interventional, non-comparative study

Assessment of the commitment level of the participants in the intervention

N = 457

(19 clusters, office workers)

No CG

6 months

Learning theory

37. Lee SH. The Effects of a Mobile Wellness Intervention with Fitbit Use and Goal Setting for Workers. 2019

Individualized behavioral program with goal setting up and counseling sessions with the use of a pedometer and motivational messages

Interventional RCT study. Incentive program (mobile wellness intervention using Fitbit, goal setting, brief counseling and motivational text messaging for workers) vs use of a pedometer alone

N = 82 (two workplaces, plant workers)

IG: N = 41

CG: N = 41

12 weeks

Self-determination continuum theory

38. Brunet J. Motivation Predicts Change in Nurses’ PA Levels During a Web-Based Worksite Intervention: Results From a Randomized Trial. 2020

Nested study, ref n°33 (Reed JL, 2018)

Originality: this second study examined mediators/moderators (such as commitment, self-efficacy and intention) of the outcomes of PA and the dose relation of goal setting with PA

N = 76

(care setting: nurses)

6 weeks

Self-determination continuum theory

39. Murray JM

Mechanisms of PA behavior change in an incentive-based intervention: mediation analysis. 2020

The PAL Scheme program integrated a novel PA remote tracking system with web-based monitoring and evidence-based behaviour change tools (i.e. self-monitoring, goal-setting)

Participants were encouraged by financial incentives to undertake 150 min/week of PA which is in line with current guidelines

(public sector, office-based employees)

N = 853

IG = 457

CG = 396

6 months

Learning theory

40. Thøgersen-Ntoumani C

Feasibility and preliminary effects of a peer-led motivationally embellished

workplace walking intervention: A pilot cluster

randomized trial (the START trial). 2020

Individualized behavioral program with goal setting up (3000 steps i.e., equivalent to a 30 mn moderate intensity walk) and counseling sessions with the use of a pedometer and motivational messages

Incentive program: all participants were provided with a Fitbit Zip and information on health benefits of walking. IG had access to a mobile phone app incorporating behavior change techniques and had a peer leader trained in a motivationally supportive communication style

N = 97 (8 worksites, office workers)

IG: N = 50

CG: N = 47

16 weeks

Self-determination continuum theory

41. Metcalfe RS. Time-efficient and computer-guided sprint interval exercise training for improving health in the workplace: a randomised

mixed-methods feasibility study in office-based employees. 2020

Mixed methods, quantitative and qualitative, to investigate the feasibility, acceptability and effectiveness of a short duration, High-intensity Exercise Intervention (REHIT) when applied unsupervised in a workplace setting

IG completed three exercise sessions per week. The 10-min exercise sessions consisted of low-intensity cycling (60 W) and one (first session) or two (all other sessions) brief'all-out'sprints (10 s. in week 1; 15 s. in weeks 2–3 and 20 s. in the final 3 weeks). The primary outcome was the change in maximal aerobic capacity (V̇O2 max). Secondary outcomes were adherence and acceptability

N = 29

(office-workers)

IG = 16

CG = 13

6 weeks

No theoretical framework

42. Welsh A. Process evaluation of a workplace-based health promotion and exercise cluster randomised

trial to increase productivity

and reduce neck pain in office workers: a RE-AIM approach. 2020

RE-AIM (reach, effectiveness, adoption, implementation, maintenance) framework:

randomised trial comparing an “ergonomic plus exercise intervention” to an “ergonomic plus health promotion intervention”

Qualitative and quantitative approach. The EET group (IG) received strength

training for 20 min, three days each week (one supervised, two unsupervised) for 12 weeks, while the EHP (CG) group received a one hour facilitated health promotion

information session each week for 12 weeks

N = 753

(industry network, public/private sector, government-operated businesses and one university, 14 organizations, office personnels, 19% were managers)

IG = 367

CG = 373

12 weeks + 12-month follow up

No theoretical framework

43. Fernandez La Puente de Battre MD. What it takes to recruit 77 subjects for a one-year study on active commuting. 2020

Motivational program with the use of financial incentives: 50- euro vouchers for bicycle shop, 24-h public transport tickets

GISMO study, randomized controlled monocentric trial

N = 73

(hospital)

IG: N = 51

CG: N = 22

53 days

No theoretical framework

44. Morris A. Sit Less and Move More—A multi-component intervention with and without height-adjustable workstations in contact centre call agents: a pilot RCT. 2021

Multi-component intervention to sit less and move more, with (SLAMM +) and without (SLAMM) height-adjustable workstations, in contact centre call agents

After randomisation, a height-adjustable workstation to individual desk was provided for SLAMM + agents (IG). Mixed-methods assessed response, recruitment, retention, attrition and completion rates, adverse effects, trial feasibility, acceptability, and effectiveness on worktime sitting

N = 59

(call centers)

IG = 30

CG = 29

10 months

Behavior change wheel

45. Tosta Maciel RRB. Does tutor’s support contribute to a telehealth program that aims to promote the quality of life of office workers? A cluster randomized controlled trial. 2021

Motivational program

The objective is to evaluate the effectiveness of a telehealth program in the workplace that was made available in two ways: a conventional telehealth program (CG) and a telehealth care program with tutors: nutritionists, psychologist and physiotherapists (IG)

The communication instrument was Moodle. Nine audiovisual sessions were performed and addressed to the all participants: 1) musculoskeletal health (walking program, back school, muscle relaxation, work related musculoskeletal diseases), 2) healthy diet (eating and commensality, ultra-processed food and food labeling, oils and fats), 3) mental health (meaning of work and burn out)

N = 326

(18 clusters, office workers, computer users from a public university)

IG = 178

CG = 148

6 months

e-Health education program, based on “socio-interactionist theory”

46. Althammer SE. Comparing Web-Based and Blended Training for Coping With Challenges of Flexible Work Designs: Randomized Controlled Trial. 2023

Flexible work designs (FWDs), such as flextime, telework, and mobile work, provide workers with temporal and spatial flexibility. This study hypothesized that a blended intervention, combining web-based self-training and face-to-face elements should increase social support and adherence compared with web-based interventions

3-armed randomized controlled trial with 2 IGs and a waitlist CG

This study also evaluated the effectiveness of an intervention in improving the recovery, work-life balance, and well-being of workers with flexible work designs (FWDs) compared with a waitlist CG

N = 373

(researchers, doctorate, technical college, master, bachelors’degree). Having a leadership position: 25% (94/373 at T2 = 4 weeks post-intervention)

IG: N = 194

IG1: allocated to web-based intervention: N = 198

IG2: allocated to blended intervention: N = 196

CG: N = 179

4 weeks

self-determination theory