Journal Logo


Optic Flow Influences Perceived Exertion and Distance Estimation but not Running Pace


Author Information
Medicine & Science in Sports & Exercise: August 2014 - Volume 46 - Issue 8 - p 1658-1665
doi: 10.1249/MSS.0000000000000257
  • Free


Most pacing control theories center around the interpretation of RPE at any moment in the exercise bout and whether such levels or rates of change in RPE can be sustained for the remaining duration of the event (34,35,38). A linear increase in RPE with time has been observed during constant load tasks (13,26,27) as well as self-paced endurance events (28,41). The increase in RPE during constant load trials has also been shown to exhibit a scalar relationship with exercise duration so that when RPE is plotted against percentage of exercise time completed, the absolute value of RPE is virtually identical at any given point through each trial, even when the conditions in which the exercise is undertaken mean that time to exhaustion is altered (26).

Recently, several slightly different explanations of how RPE is interpreted to regulate effort during exercise have been proposed, which involve either template matching (38), duration-based risk evaluation (11), estimates of time to fatigue (9,15,16), or rate of increase in RPE (26). In the RPE template model, pace is determined according to how well experienced levels of RPE compare against expected RPE at any given point in the exercise bout (38), with any deviation resulting in an adjustment in pace to bring RPE back to the expected value. Although still a matter of some debate (23), this model assumes that RPE is generated as a result of afferent information about the internal condition of the body and so serves as a mechanism to protect the vital functioning of the body that may be compromised if homeostasis is threatened. Evidence for the RPE template model is mainly derived from studies where deception was used to create a mismatch between actual and expected exercise duration, resulting in an altered RPE response and incorrect control of pacing (1–3).

Hazard score has been put forward as an alternative explanation of pacing control that places more emphasis on duration-based risk evaluation, whereby athletes’ pacing judgments result from a calculation of the product of momentary RPE and the proportion of the task remaining (11).

A further conceptualization of how athletes use RPE to control their pace centers around evidence that, at a given a given work rate, athletes are able to predict their time to exhaustion (9,15,16,17), and this estimated time limit to exhaustion is an important component in any calculation as to the most appropriate pace throughout the exercise bout.

The rate of increase in RPE value is suggested to result primarily from peripheral factors via afferent feedback and, additionally, central factors such as mental fatigue as well as environmental factors in the flush model (25). In this model, the rate of increase in RPE is dependent on both this combined feedback and the total central capacity for accumulating this feedback.

Common to each of the models described earlier is an awareness of the event end point that is incorporated into the calculations used to set pace (9,11,33,39). Several studies have used false feedback about speed, distance, or time and found an effect on RPE, pacing, and performance, albeit of different magnitudes (1–3,24). There is also growing evidence that exteroceptive cues, such as visual sensation or optic flow, have similar effects (27).

Optic flow is the sense of movement through an environment caused by the expanding patterns of flow on the retina (18). Treadmill walking in the absence of optic flow has been shown to lead to a perceptual aftereffect, whereby participants judge distances to be further than if they had experienced optic flow while walking (30). Treadmill walking in the presence of optic flow slower than actual walking speed results in subsequent overshoot when subsequently asked to walk a specific distance and undershoot when the treadmill walking was conducted in the presence of optic flow faster than treadmill speed (32). Studies examining the effect of prior exercise with optic flow of various speeds have shown a recalibration of visuomotor control (12). More recently, the manipulation of optic flow during constant workload cycling was found to influence RPE, whereby slower optic flow induced both lower overall levels of RPE as well as a shallower rate of increase in RPE (27). Given the fairly well-established role of RPE in controlling pace, what is not yet known is whether optic flow–induced variations in RPE could lead to altered pacing.

The purpose of this study was to investigate whether altered rates of optic flow result in an altered ratio of perceived to actual running distance and whether such an altered perception of the ratio of speed to work would result in modified pacing behavior.



Twelve moderately trained competitive male athletes participated in this study whose mean ± 1 SD age, stature, and body mass was 20.1 ± 1.4 yr, 180.2 ± 7.6 cm, and 74.5 ± 9.4 kg. Participants were active in varsity sport, competing in either association football or squash, and completed 5.25 ± 0.87 training sessions per week, for a total of 8.4 ± 2.8 h each week. Each participant provided written informed consent to take part in this study, which was approved by the ethics committee of the University of Essex.


A two-way repeated-measures experimental design was used in which participants performed a familiarization self-paced 5-km running time trial on a treadmill followed by three 5-km running self-paced trials under different optic flow conditions (condition factor). During each of the running tasks, participants were not given any feedback as to the distance or time elapsed. Participants were asked to indicate when they perceived that they had completed each 1-km segment of the run by calling out their RPE at that point. Elapsed time and actual distance were recorded at the moments when participants declared reaching each 1-km segment (perceived distance factor) and when they actually reached each 1-km segment (actual distance factor). Participants were randomly assigned to complete the three optic flow running trials in a counterbalanced order. Each participant performed all running trials at the same time of day (±1 h), with a recovery interval varying between 3 and 7 d.

Treadmill running time trial procedures.

Participants completed all running trials on a Powerjog JW200 motorized treadmill (Sport Engineering Ltd, Birmingham, UK). In the optic flow trials, participants watched a video footage of moving along a path through a forest, as would be seen from the point of view of a runner. The video footage was projected on to a large screen, 2.2 × 2.2 m in size, 3 m in front of the treadmill. The speed of the video footage was linked to the speed that the participant was running via a foot pod that each participant wore on their shoe, which controlled the video via Virtual Runner software (Outside Interactive, South Easton, MA). The room in which the trials took place was darkened, and the treadmill was flanked by screens to obscure the view of the rest of the room, with the intention of directing the attention of the participant toward the video footage. Other than that generated by the treadmill, no other noise was audible during the trials. The temperature of the room was controlled at a temperature of 19°C. Investigators provided no verbal encouragement or instruction during the trials and remained out of sight of the participant.

Each participant first performed a self-paced 5-km familiarization time trial without video footage, in which they were instructed to complete in the fastest possible time. Participants then performed three 5-km running trials under different optic flow conditions (described in the next section), in which they were also asked to complete in the fastest possible time. During each trial, participants were free to control the speed of the treadmill by using the speed increase and decrease buttons. All information on the control panel of the treadmill relating to actual speed and time was obscured from the participants’ view. Participants were asked to refrain from eating solid food or consuming caffeine before each running trial for 2 and 4 h, respectively. Before each running trial, participants performed a 5-min self-paced warm-up with a 5-min rest interval.

Optic flow simulation.

During the freely paced running trials, subjects were instructed to observe a video footage of a real path through a forest being traveled, which was projected onto a large screen in front of them. Unknown to the participants, the speed of the video footage (optic flow) was varied in each condition so that it matched their true running speed (RNORM), was 25% slower than their true running speed (RSLOW), or was 25% faster than their true running speed (RFAST). Variation in optic flow was achieved by altering the calibration distance within the Virtual Runner software, in which the software uses to alter the relationship between the speed indicated by the foot pod and the rate of forward velocity of the video footage. As previously described, participants were randomly allocated to perform the three running tasks in a counterbalanced order of optic flow conditions. During a postexperimental debrief, some participants reported that they were aware that there was something different between trials, but none were able to accurately identify the nature of the manipulations that had taken place.

Perceived exertion measurements.

During each running trial, participants were asked to provide an overall RPE each time they perceived that they had covered a 1-km segment of the 5-km trial using the Borg 6-20 RPE scale (6). Each subject was familiarized with the RPE scale, which was administered in accordance with published standardized instructions (6). Each trial continued until the participant had provided their final RPE score, indicating that they perceived that they had run 5 km. If the participant provided their five 1-km scores before the completion of the full 5-km actual distance, the participant was instructed to continue running until it was indicated to them that they had completed the full distance.

Statistical analysis.

A two-way repeated-measures ANOVA was used to analyze condition-by-distance differences in perceived distance and running time (performance). Errors in estimated running distance were calculated as the percentage deviation of the actual distance covered by participants compared with each kilometer point. For example, a participant who stated they had reached 1 km but had actually run 1.1 km would have a +10% distance estimation error. Because of RPE only being recorded at the perceived kilometer points, RPE scores were normalized by expressing them as the actual distance covered divided by the RPE, which gives a measure of the distance covered per RPE point on the Borg 15-item scale. The normalized data were then reanalyzed using a two-way repeated-measures ANOVA. The second approach involved conducting a two-way repeated-measures ANCOVA using running speed for each 1-km segment as a covariate. Differences in running speed between optic flow conditions were too small for running speed to be treated as a covariate, and so these results are not presented. Pearson’s product moment correlation tests were used to measure relationships between RPE and actual running distance and perceived running distance. All results are expressed as mean values with one SD and effect sizes as partial eta squared (ηp2) or eta squared (η2). An alpha level of 0.05 was used to indicate statistical significance.



The average overall completion time (min:s) for the self-paced 5-km familiarization time trial was 1452 ± 192 s (min:s, 24:12 ± 3:12). The average completion times for the self-paced run trials were RSLOW 1477 ± 138 s (min:s, 24:37 ± 2:18), RNORM 1514 s ± 179 s (min:s, 25:14 ± 2:59), and RFAST 1524 ± 189 s (min:s, 25:24 ± 3:09). A one-way within-subjects ANOVA showed no difference in the 5-km completion times between conditions ([Latin Small Letter Open E] = 0.99, F2,22 = 1.8, P = 0.189, ηp2 = 0.14).


A two-way within-subjects ANOVA showed no condition-by-distance interaction for the pacing of actual running distance (F2.9,31.5 = 2.6, P = 0.071, ηp2 = 0.19), no condition main effect (F2,22 = 1.5, P = 0.25, ηp2 = 0.12), and no distance main effect (F1.5,16.2 = 1.7, P = 0.220, ηp2 = 0.13). The condition-by-distance outcomes for the actual run pace at each kilometer run point are presented in Figure 1A.

Condition-by-distance outcomes for actual running pace (A), perceived running distance (B), and distance estimation error expressed in kilometers (C) and as a percentage (D).

A two-way within-subjects ANOVA showed a condition-by-distance interaction for the pacing of perceived running distance (F3.4,37.3 = 3.6, P = 0.001, ηp2 = 0.25), a condition main effect (F2,22 = 12.6, P < 0.001, ηp2 = 0.54), and a distance main effect (F4,44 = 3.9, P = 0.002, ηp2 = 0.32). The condition-by-distance pacing outcomes for perceived running distance are presented in Figure 1B.

Errors in estimated running distance.

A two-way within-subjects ANOVA showed a condition-by-distance interaction for the cumulative actual distance at perceived kilometer points (F2.7,29.3 = 16.5, P < 0.001, ηp2 = 0.60), a condition main effect (F2,22 = 15.0, P < 0.001, ηp2 = 0.58), and a distance main effect (F1.3,14.2 = 724.1, P < 0.001, ηp2 = 0.99).

A two-way within-subjects ANOVA showed no condition-by-distance interaction for the percentage error in perceived distance (F2.8,31.3 = 2.8, P = 0.058, ηp2 = 0.20) and no distance main effect (F1.2,13.7 = 4.2, P = 0.053, ηp2 = 0.28), but there was a condition main effect (F2,22 = 10.6, P = 0.001, ηp2 = 0.49). The condition-by-distance outcomes for distance run at each perceived kilometer run point and for the percentage error in perceived distance are presented in Figures 1C and 1D.

Perceived exertion.

Pearson’s product moment correlation tests showed a significant, positive correlation between actual running distance and RPE during RNORM (r = 0.990, P = 0.001), RFAST (r = 0.996, P < 0.001), and RSLOW (r = 0.995, P < 0.001). There were also positive correlations between perceived running distance and RPE in RNORM (r = 0.990, P = 0.001), RFAST (r = 0.993, P = 0.001), and RSLOW (r = 0.996, P < 0.001). RPE against actual and perceived running distance is presented in Figures 2A and 2B.

Scattergram showing the relationship between RPE and actual running distance (A) and perceived running distance (B). Condition-by-distance interaction for running distance/RPE (C).

Perceived exertion against perceived run distance.

A two-way within-subjects ANOVA showed no condition-by-perceived distance interaction for RPE (F2.6,28.6 = 1.1, P = 0.384, ηp2 = 0.089) and no condition main effect (F2,22 = 2.3, P = 0.128, ηp2 = 0.17), but there was a significant distance main effect (F1.2,12.7 = 57.5, P < 0.001, ηp2 = 0.84). However, because RPE was self-reported at the perceived kilometer points, the actual distance that participants ran between providing RPE values varied. To normalize the work completed between RPE scores, the actual distance completed at each perceived kilometer point was divided by the RPE score, giving values of the actual distance covered per RPE point on the scale as an indication of the actual work that is required to be completed to elicit an increase in the RPE score in each condition. A two-way within-subjects ANOVA showed a condition-by-distance interaction for the actual distance divided by RPE (F2.7,29.9 = 5.3, P < 0.001, ηp2 = 0.33), a condition main effect (F2,22 = 8.0, P = 0.002, ηp2 = 0.42), and a distance main effect (F1.2,13.2 = 180.4, P < 0.001, ηp2 = 0.94). The condition-by-distance outcomes for running distance/RPE are presented in Figure 2C.

Hazard scores.

Hazard scores for actual run distance were calculated by multiplying the RPE score given at each perceived kilometer point by the proportion of the actual 5-km run distance remaining at that point. A two-way within-subjects ANOVA showed a condition-by-distance interaction for hazard score (F2.5,27.4 = 15.6, P < 0.001, ηp2 = 0.59), a condition main effect (F2,22 = 15.7, P < 0.001, ηp2 = 0.59), and a distance main effect (F1.2,13.7 = 149.7, P < 0.001, ηp2 = 0.93). The condition-by-distance outcomes for actual hazard score at each perceived kilometer run point are presented in Figure 3A.

Condition-by-distance interactions for hazard score calculated using actual running distance (A) and hazard score calculated using perceived running distance (B). Note: Theoretically, hazard scores should never fall below zero, but in this instance, this occurred because the actual running distance after 5 km exceeded perceived duration of 5 km.

Hazard scores for perceived run distance were calculated by multiplying the RPE score at each perceived kilometer point by the proportion of the perceived 5-km run distance remaining at that point. A two-way within-subjects ANOVA showed no condition-by-distance interaction for hazard score (F3.7,40.8 = 0.537, P = 0.697, ηp2 = 0.047) and no condition main effect (F2,22 = 0.965, P = 0.397, ηp2 = 0.081), but there was a distance main effect (F1.2,13.5 = 326.7, P < 0.001, ηp2 = 0.97). The condition-by-distance outcomes for perceived hazard score at each perceived kilometer run point are presented in Figure 3B.


The main finding of this study is that different rates of optic flow lead to an altered ratio of perceived to actual distance during running, with slow optic flow resulting in an elongated perception of distance and fast optic flow leading to a compressed perception of distance relative to actual running distance. This finding is consistent with several studies that have reported residual effects of incongruent optic flow during exercise on postexercise distance estimation (30,32). One outcome of the altered ratio of perceived to actual distance was that participants’ actual run distance completed at the point they believed they had completed the full 5 km was, on average, 423.5 m shorter than 5 km in RFAST, 388 m further than 5 km in RNORM, and 1067 m further than 5 km in RSLOW. This amounts to an undetected variation in running distance of between −8.5% and +21.3%. Another key finding of this study, which corroborates our previous findings (28), is the lower RPE that occurred during conditions of slow optic flow. What is new is that pacing and performance were unaffected by slow optic flow, or the lower RPE values observed in this condition. This finding, which suggests factors other than RPE may influence changes in pace, is somewhat surprising given the hitherto robustness of the RPE template model (38).

The RPE template model (38) predicts that misinformation about exercise duration would result in a change of pace. Our results are contrary to the predictions of the RPE template model in that participants did not adopt a behavioral response (either a modulated pace throughout the trial or a reduced pace when the anticipated duration was exceeded) to deal with the apparent mismatch between momentary RPE and estimation of remaining duration. Rather, an alternative, possible explanation is that participants used a cognitive strategy whereby the scalar relationship between rate of increase in RPE and time was apparently altered, a pacing control mechanism that is not accounted for in the RPE template model (38). The modification of the scalar relationship between RPE and time can in fact account for the findings of several studies where familiarization/experience time trials, in which participants are deceived about their performance, lead to modified pacing or performance in subsequent time trials despite altered RPE (24,29,36). Put another way, the scalar relationship between RPE and time or distance are modified with experience, which influences the pacing and performance of subsequent trials.

Hazard scores.

Hazard scores, where the momentary RPE is multiplied by the proportion of total distance of the task remaining, is suggested to represent the “language” by which anticipatory pacing can take place (11). In the present study, optic flow is suggested to provide feedback about the running speed, which when coupled with an estimation of elapsed time provides information about distance covered and therefore proportion of the task remaining. As participants were not provided with information about elapsed time, there is potential for error in the estimation of time, which coupled with the deceptive optic flow presents two variables through which error in the estimation of the remaining task can occur. Nonetheless, it would be the perception of distance covered that participants would use to calculate the hazard score at any point during the task. Using perceived distance results in identical hazard score curves during the task for the different optic flow conditions, which should result in identical pacing, which is what was observed. It is theorized that the high hazard score represents a signal to slow down, the moderate hazard score represents a signal to maintain pace, and the low hazard score represents an inducement to increase pace. However, an analysis of the actual pacing shows that there was no distance main effect such that there was no apparent end spurt, so the perceived hazard score is not consistent with the finding that there was no alteration in pacing with increasing distance, as would be predicted by a reduction in hazard score and an apparent inducement to accelerate as the task nears completion.

In contrast to the identical actual pacing, perceived pacing shows different profiles for the three conditions, indicating that the hazard score calculation based on perceived distance may not have been influencing their beliefs about their pace. Participants will have been consciously aware of their actions in controlling the speed of the treadmill, and yet significant differences in perceived pace were displayed in the first and the fifth kilometers between the fast and the slow conditions, with the participants in the fast condition perceiving that they had started slowly and increased their pace progressively throughout the 5 km. However, in the slow condition, they believed their pace to be fairly even throughout. This potentially presents a contradiction between, on the one hand, the hazard scores based on perceived distance and, on the other hand, the calculated hazard scores being inconsistent with the perceived pacing of participants. This may present a problem for the explanation of the cognitive mechanism by which the hazard score may be used in pacing control. This lack of correspondence between perception and action has previously been illustrated in a study where participants inaccurately visually judged distances between objects but were able to accurately judge the correct distance to walk between them when blindfolded. Also, when walking blindfolded, parallel to the objects, they were able to accurately point to where they estimated the objects would be (22). Several studies in which the overestimation of the gradient was not matched by a visually guided action exist, which was more consistent with the accurate assessment of steepness of gradients (4,10,31). This apparent separation of action and perception is further supported by evidence that visual information dealing with, on the one hand, object identification and, on the other hand, visually guided motor actions is dealt with via two separate neural pathways (19,40,42). Hence, the control of motor action during movement and the assessment of how difficult an action will be are separate cognitive processes. Indeed, it has been suggested that physical movement is controlled by the subconscious and that the conscious sense of volitional control is, in fact, illusory and is generated after the initiation of movement has already occurred (21). In this way, conscious perception, which may be subject to modification by cognitive appraisal, for example, may become divergent from action. A similar explanation is provided by circumstances where proprioceptive and visual information are in conflict, resulting in an effect known as visual capture, where beliefs about motor action are biased toward visual information (20). The previous explanations, alongside our findings, may potentially raise doubts about the importance of consciously experienced RPE in the control of pacing.

Estimated time limit.

In a recent review (9), the ability of athletes to fairly accurately predict their time to exhaustion was suggested to fit well with a psychobiological model of exercise performance (23). In this model, previous experience of perceived exertion during exercise bouts of varying intensity and duration plays a key role in the modulation of pace, and it has been suggested that the estimation of time to exhaustion provides information about this component of the model (16). Our findings suggest that estimated time limit may be subject to modification by optic flow. Because the rate of increase in RPE is set in memory, a possible mechanism could be via an altered sense of the passage of time to maintain a scalar relationship between RPE increase and exercise duration (14,26). Figure 2B illustrates that the mean magnitude of change in the ratio of perceived to actual distance at the 5-km point relative to RNORM was +12.6% in RSLOW and −15.1% in RFAST, and Figure 2C shows that normalized RPE (d/RPE) at the 5-km point relative to RNORM was + 11.2% in RSLOW and −11.8% in RFAST, indicating very similar magnitudes of effect and perhaps providing an indication that the altered perception of distance is the likely mechanism involved.

Schoolchildren performing a self-paced running task have been shown to be better able to judge the extent of the task and produce better performances when instructed to complete a given distance rather than a given time (8). Time is an abstract mental concept rather than a visual extent that can be directly perceived. When the anticipated temporal–spatial relationship is distorted, it could be that the visual extent takes precedence. Indeed, the perception of time and space are thought to be highly interdependent with recent work, indicating that perceptions of duration are susceptible to alteration by changes in the visual impressions of spatial displacement, whereas changes in duration do not affect the perception of spatial displacement (7). In the present study, subjects were deceived by the altered optic flow as to the distance they had actually covered in a given estimated time, and if they were to complete the 5 km with the same scalar relationship between RPE increase and duration or distance that they had previously experienced during the familiarization trial, then they had to either reject the feedback they were receiving about spatial extent or alter their perception of the passage of time. Failure to do one of these two things would have resulted in a change in the scalar relationship between the rate of RPE increase and the duration or distance. Figure 4 illustrates the interrelated nature of work rate, passage of time, and rate of RPE increase.

The interrelated nature of work rate, passage of time, and rate of increase in RPE.

One limitation of the present study is that participants had to make a decision to alter their pace about which they would be consciously aware, which involved physically altering their speed via the controls on the side of the treadmill. Much neurological evidence exists to suggest that decisions to make voluntary movements are taken before the information about that decision reaches consciousness (21), and as such, the involvement of executive decision making, associated with consciousness, may have distorted the dynamic relationship between work rate, temporal–spatial ratio, and scalar increase in RPE. In freely paced running, some decisions about pacing may not have emerged into consciousness, with more immediate changes in pace taking place and shifting the response in the RPE template away from altered perception of time, as occurred in the present study. Such subconscious pace adjustments have previously been demonstrated during both cycling and running (5,37). The requirement for executive decision making involvement in pacing may also have further complicated the relationship between homeostatic control and motor action and so may therefore have led to suboptimal performances. One other confounding factor for participants was the uncertainty about the end point of the exercise bout, as the possibility always existed that they may have to continue their exercise beyond their anticipated end point if they had underestimated the actual 5-km distance. This is likely to have suppressed any tendency to engage in an end spurt and may also have increased their tendency to withhold some energy reserve as contingency in the event of having to continue further than anticipated, also potentially leading to submaximal performances. Although much effort was made to reduce the possibility for participants to be influenced by unintended sensory signals, one potential indicator of speed could have been the sound made by the treadmill at different work rates, which was not possible to eliminate.


Optic flow during exercise alters the ratio of perception of distance to actual distance covered. In the present study, this did not lead to altered pacing behavior, but instead, participants seemed to adopt a cognitive strategy that maintained the scalar relationship between the rate of increase in RPE and the perceived duration of the task. Supporting the findings of our previous research (28), the maintenance of the scalar RPE increase resulted in an altered work–RPE ratio in different optic flow conditions. Two cognitive mechanisms are put forward to account for the altered work–RPE ratio. The first is the elicitation of associative emotional responses based on previous experience of running bouts where certain rates of optic flow are recalled to have been experienced, which are associated with the conscious experience of particular RPE scores. Alternatively, participants are basing their perceptual experiences on a belief about the rate at which RPE should increase during particular perceived exercise intensities, and the altered optic flow rates challenge this belief, leading to an altered perception of the passage of time, which maintains the scalar relationship of RPE with exercise duration. The failure of participants to alter their pace in accordance with the revised anticipated duration of the task suggested by the optic flow or to significantly slow down once the preexercise anticipated exercise duration was exceeded seems to contradict the predictions made by the RPE template model (3). Participants’ perceived hazard scores were identical in all optic flow conditions, which is consistent with the finding that actual pacing was also identical. However, participants’ perceived pacing was significantly different between conditions, suggesting a divergence between perceived and actual actions that calls into question the mechanism of the hazard score and warrants further investigation.

This study was funded internally by the University of Essex, and no other source of external funding or support was used. The authors acknowledge Ben Wakile, Daniel Byron, and Rowan Cooke for their help with data collection. None of the authors have professional relationships with companies or manufacturers who will benefit from the results of this study.

None of the results presented in this study constitute endorsement by the American College of Sports Medicine.


1. Albertus Y, Tucker R, Gibson ASC, Lambert EV, Hampson DB, Noakes TD. Effect of distance feedback on pacing strategy and perceived exertion during cycling. Med Sci Sports Exerc. 2005; 37 (3): 461–8.
2. Ansley L, Robson PJ, St Clair Gibson A, Noakes TD. Anticipatory pacing strategies during supramaximal exercise lasting longer than 30 s. Med Sci Sports Exerc. 2004; 36 (2): 309–14.
3. Baden DA, McLean TL, Tucker R, Noakes TD, St Clair Gibson A. Effect of anticipation during unknown or unexpected exercise duration on rating of perceived exertion, affect, and physiological function. Br J Sports Med. 2005; 39 (10): 742–6.
4. Bhalla M, Proffitt DR. Visual–motor recalibration in geographical slant perception. J Exp Psychol Hum Percept Perform. 1999; 25 (4): 1076–96.
5. Billat VL, Wesfreid E, Kapfer C, Koralsztein JP, Meyer Y. Nonlinear dynamics of heart rate and oxygen uptake in exhaustive 10,000 m runs: influence of constant vs. freely paced. J Physiol Sci. 2006; 56 (1): 103–11.
6. Borg GA. Borg’s Perceived Exertion and Pain Scales. Champaign (IL): Human Kinetics; 1998. pp. 44–9.
7. Casasanto D, Boroditsky L. Time in the mind: using space to think about time. Cognition. 2008; 106 (2): 579–93.
8. Chinnasamy C, St Clair Gibson A, Micklewright D. Effect of spatial and temporal cues on athletic pacing in schoolchildren. Med Sci Sports Exerc. 2013; 45 (2): 395–402.
9. Coquart JB, Eston RG, Noakes TD, et al. Estimated time limit: a brief review of a perceptually based scale. Sports Med. 2012; 42 (10): 845–55.
10. Creem SH, Proffitt DR. Two memories for geographical slant: separation and interdependence of action and awareness. Psychon Bull Rev. 1998; 5 (1): 22–36.
11. de Koning JJ, Foster C, Bakkum A, et al. Regulation of pacing strategy during athletic competition. PloS One. 2011; 6 (1): e15863.
12. Durgin FH, Pelah A. Visuomotor adaptation without vision? Exp Brain Res. 1999; 127 (1): 12–8.
13. Eston R, Faulkner J, St Clair Gibson A, Noakes T, Parfitt G. The effect of antecedent fatiguing activity on the relationship between perceived exertion and physiological activity during a constant load exercise task. Psychophysiology. 2007; 44 (5): 779–86.
14. Faulkner J, Parfitt G, Eston R. The rating of perceived exertion during competitive running scales with time. Psychophysiology. 2008; 45: 1077–85.
15. Garcin M, Coquart JB, Robin S, Matran R. Prediction of time to exhaustion in competitive cyclists from a perceptually based scale. J Strength Cond Res. 2011; 25 (5): 1393–9.
16. Garcin M, Coquart J, Salleron J, Voy N, Matran R. Self-regulation of exercise intensity by estimated time limit scale. Eur J Appl Physiol. 2012; 112 (6): 2303–12.
17. Garcin M, Vautier JF, Vandewalle H, Wolff M, Monod H. Ratings of perceived exertion (RPE) during cycling exercises at constant power output. Ergonomics. 1998; 41 (10): 1500–9.
18. Gibson JJ. The Perception of the Visual World. Boston: Houghton Mifflin; 1950. pp. 131–6.
19. Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci. 1992; 15 (1): 20–5.
20. Guarraz M, Provost S, Narison R, Brugnon A, Virolle S, Bresciani J-P. Integration of visual and proprioceptive afferents kinesthesia. Neuroscience. 2012; 223: 258–68
21. Hallett M. Volitional control of movement: the physiology of free will. Clin Neurophysiol. 2007; 118 (6): 1179–92.
22. Loomis JM, Da Silva JA, Fujita N, Fukusima SS. Visual space perception and visually directed action. J Exp Psychol Hum Percept Perform. 1992; 18 (4): 906–21.
23. Marcora S. Counterpoint: afferent feedback from fatigued locomotor muscles is not an important determinant of endurance exercise performance. J Appl Physiol. 2010; 108 (2): 454–7.
24. Micklewright D, Papadopoulou E, Swart J, Noakes T. Previous experience influences pacing during 20-km time trial cycling. Br J Sports Med. 2010; 44: 952–60.
25. Millet GY. Can neuromuscular fatigue explain running strategies and performance in ultra-marathons? Sports Med. 2011; 41 (6): 489–506.
26. Noakes TD. Linear relationship between the perception of effort and the duration of constant load exercise that remains. J Appl Physiol. 2004; 96 (4): 1571–2; author reply 1572-3.
27. Parry D, Chinnasamy C, Papadopoulou E, Noakes T, Micklewright D. Cognition and performance: anxiety, mood and perceived exertion among Ironman triathletes. Br J Sports Med. 2011; 45 (14): 1088–94.
28. Parry D, Chinnasamy C, Micklewright D. Optic flow influences perceived exertion during cycling. J Sport Exerc Psychol. 2012; 34 (4): 444–56.
29. Paterson S, Marino FE. Effect of deception of distance on prolonged cycling performance. Percept Mot Skills. 2004; 98 (3): 1017–26.
30. Proffitt DR, Stefanucci J, Banton T, Epstein W. The role of effort in perceiving distance. Psychol Sci. 2003; 14 (2): 106–12.
31. Proffitt DR, Bhalla M, Gossweiler R, Midgett J. Perceiving geographical slant. Psychonomic Bull. Rev. 1995; 2 (4): 409–28.
32. Reiser JJ, Pick HL, Ashmead DH, Garing AE. Calibration of human locomotion and models of perceptual-motor organization. J Exp Psychol Hum Percept Perform. 1995; 21: 480–97.
33. St Clair Gibson A, Baden DA, Lambert MI, et al. The conscious perception of the sensation of fatigue. Sports Med. 2003; 33 (3): 167–76.
34. St Clair Gibson A, Noakes TD. Evidence for complex system integration and dynamic neural regulation of skeletal muscle recruitment during exercise in humans. Br J Sports Med. 2004; 38 (6): 797–806.
35. St Clair Gibson A, Lambert EV, Rauch LHG, et al. The role of information processing between the brain and peripheral physiological systems in pacing and perception of effort. Sports Med. 2006; 36 (8): 705–22.
36. Stone MR, Thomas K, Wilkinson M, Jones AM, St Clair Gibson A, Thompson KG. Effects of deception on exercise performance: implications for determinants of fatigue in humans. Med Sci Sports Exerc. 2012; 44: (3): 534–41.
37. Terblanche E, Wessels JA, Stewart RI, Koeslag JH. A computer simulation of free-range exercise in the laboratory. J Appl Physiol. 1999; 87 (4): 1386–91.
38. Tucker R. The anticipatory regulation of performance: the physiological basis for pacing strategies and the development of a perception-based model for exercise performance. Br J Sports Med. 2009; 43 (6): 392–400.
39. Ulmer H-V. Concept of an extracellular regulation of muscular metabolic rate during heavy exercise in humans by psychophysiological feedback. Experimentia. 1996; 52 (5): 416–20.
40. Ungerleider LG, Mishkin M. Object vision and spatial vision: two cortical pathways. Trends Neurosci. 1983; 10: 414–7.
41. Utter AC, Kang J, Nieman DC, et al. Ratings of perceived exertion throughout an ultramarathon during carbohydrate ingestion. Percept Mot Skills. 2003; 97 (1): 175–84.
42. Witt JK, Proffitt DR. Perceived slant: a dissociation between perception and action. Perception. 2007; 36 (2): 249–57.


© 2014 American College of Sports Medicine