Recalibrating vision-for-action requires years after sight restoration from congenital cataracts
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This paper will be of interest to researchers in the fields of motor control, visual perception, learning and brain plasticity, sight loss and rehabilitation. The paper shows the contributions of sensory-motor experience to the development of visuo-motor recalibration abilities using careful experimental methods and analyses, comparing a rare population of late-operated cataract patients with control groups.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Neuroscience (eLife)
Abstract
Being able to perform adept goal-directed actions requires predictive, feed-forward control, including a mapping between the visually estimated target locations and the motor commands reaching for them. When the mapping is perturbed, e.g., due to muscle fatigue or optical distortions, we are quickly able to recalibrate the sensorimotor system to update this mapping. Here, we investigated whether early visual and visuomotor experience is essential for developing sensorimotor recalibration. To this end, we assessed young individuals deprived of pattern vision due to dense congenital bilateral cataracts who were surgically treated for sight restoration only years after birth. We compared their recalibration performance to such distortion to that of age-matched sighted controls. Their sensorimotor recalibration performance was impaired right after surgery. This finding cannot be explained by their still lower visual acuity alone, since blurring vision in controls to a matching degree did not lead to comparable behavior. Nevertheless, the recalibration ability of cataract-treated participants gradually improved with time after surgery. Thus, the lack of early pattern vision affects visuomotor recalibration. However, this ability is not lost but slowly develops after sight restoration, highlighting the importance of sensorimotor experience gained late in life.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
The authors asked to what extent early visual and visuomotor experience is essential for developing the ability to recalibrate the visuo-motor system flexibly. This kind of recalibration crucially underpins everyday actions, allowing the brain to issue effective feed-forward motor control commands that correctly account for temporary changes in sensory-motor mappings (e.g. when using tools, carrying objects, wearing new glasses). To address the role of experience in developing these recalibration abilities, they used the unusual clinical population of late-operated cataract patients: children and adolescents who initially had many years of sensory experience that is atypical in that it lacked effective pattern vision. They used a standard sensory-motor task in which participants point to …
Author Response
Reviewer #1 (Public Review):
The authors asked to what extent early visual and visuomotor experience is essential for developing the ability to recalibrate the visuo-motor system flexibly. This kind of recalibration crucially underpins everyday actions, allowing the brain to issue effective feed-forward motor control commands that correctly account for temporary changes in sensory-motor mappings (e.g. when using tools, carrying objects, wearing new glasses). To address the role of experience in developing these recalibration abilities, they used the unusual clinical population of late-operated cataract patients: children and adolescents who initially had many years of sensory experience that is atypical in that it lacked effective pattern vision. They used a standard sensory-motor task in which participants point to targets with and without displacement of the visual image via a prism lens: after the prism displacement, the visuo-motor mapping needs to be recalibrated to enable effective pointing. They compared late-operated cataract patients with controls matched in age, controls matched in both age and visual acuity (via added visual blur), as well as an extensive broader comparison group of typically developing 6- to 17-year-olds. Their key findings were that recalibration was less effective - both in the initial effect and in the subsequent after-effect - in the patient group than in control groups; this was not related to chronological age but was related to time post-operation, such that performance came to match controls after around 2 years of improved visual experience. The authors conclude that flexible sensory recalibration abilities normally rely on extensive sensory-motor experience in childhood, and suggest that the underlying computational problem is establishing the correct correspondences between sensory and motor coordinate frames. This may be achieved through extended exposure to the sensory consequences of self-generated movements.
Strengths of the approach include use of the established (although rare and difficult to access) model population of late-operated cataract patients and a well-established experimental task (pointing after displacement of the visual image by viewing through prism lenses). The task has a known typical time-course of behaviour - supplemented here by an extensive additional study on typical development using the exact same main task, which even alone would be a meaningful contribution to literature on sensory-motor development. The procedure, measures, analysis, and the approach to control groups are careful and rigorous. The findings are rich in showing not only an initial deficit in patient vs control groups but also an approximate time course for further learning and development after which point (by ~2 years) the patients come to match controls. A challenge is the heterogenous group, in terms of age at operation and ages at testing and follow-up. However, this is very usual and almost inevitable in the literature with this kind of population, and is dealt with well in the analyses. The approach is also well supplemented by repeated follow-up of a portion (actually more than half) of the group.
One potential issue is the role of baseline pointing precision differences across the groups. It would be useful to better understand the potential role of the reduced pointing precision that was found in the cataract group (Supplemental Figure 1B). It is not surprising that, following visual deprivation, this group's predictive feedforward visuo-motor control was less precise than that of controls, even in the baseline measures before any prism manipulation, and even when the controls' vision is comparably blurred. It seems likely (although is not shown) that during the adaptation phase and the post-adaptation phase, the variability of individuals around their (gradually shifting) mean pointing location would also be higher than in controls. I wonder how large an explanatory role there could be simply for this noisier initial visuo-motor mapping in the patient group. It might be said that, on each trial, they intend to carry out a feedforward plan with a certain endpoint, but because of noise, they are on average substantially further from that endpoint than comparable controls are. So, during recalibration, while controls are dealing mainly with cancelling out one kind of error - the constant error due to the prism adaptation - the cataract patients are also dealing with more variable errors due to their own noisier visuo-motor system. In theory, could this alone - higher initial noise in the system - explain the difference? This seems like a simpler explanation than that the system has developed differently in substantial ways to do with its abilities to learn and adapt. One starting point for checking in to this would be asking if initial pointing variability predicts recalibration (perhaps controlling for visual acuity), both at first test and in the repeated participants. Another would be looking into ways to perturb controls' baseline pointing performance further (perhaps with something like an unexpected added weight rather than more visual blurring) so that their variable pointing errors were matched to the cataract group.
We thank Reviewer 1 for drawing our attention to this important point. The Reviewer is right in suggesting that precision at baseline (measured as the variance of the pointing errors in the pre-prism phase) might predict recalibration abilities (as measured by the recalibration index irecal ). Indeed, we found that the variance of the errors in pre-prism phase correlates with irecal in cataract-treated participants. Thus, the higher sensorimotor noise in cataract-treated participants (indicating more uncertainty) slows down their rate of recalibration. This finding is in accordance with Burge and colleagues (2008) who found that higher uncertainty (in their case in the form of visual blur leading to more motor variability) slows down the adaptation rate. We have now reported this analysis in the Results section and discussed the contribution of sensorimotor noise to recalibration in the Discussion. However, higher sensorimotor noise cannot explain alone the performance of the cataract-treated individuals. Indeed, the subset of participants tested a second time after surgery (4-to-16 months after the first post-surgery test) presented better recalibration ability (i.e., higher irecal ), although their precision at baseline did not increase accordingly, but stayed basically unchanged. Moreover, in their second test, their precision at baseline did not correlate with the successive irecal.
In the Discussion, we added the greater sensorimotor noise as a factor contributing to recalibration. However, as it does not explain alone the improvement of recalibration performance over time, we still discuss the contribution of their lack of experience with the sensorimotor mapping to their recalibration performance.
Another question is how well the contrast sensitivity function (CSF) as a whole (not just the maximum acuity point) was matched - this is dealt with only briefly. I am not sure to what extent the blurring manipulation would be expected to change the shape of the CSF as a whole to be in line with that of patients, and to what extent other aspects of the CSF besides the maximum acuity point determine the precision and accuracy of ballistic pointing movements under the experimental and lighting conditions used in the study. Depending on the answers to these questions, the concern could be that visual differences relevant to control of pointing remained across the patient and blurred control groups.
We have now provided more information on this point in the Methods section and in the Supplementary Information. In a pilot study, we determined the range of distances between the blurring screen and the visual target that would be needed to reproduce–in controls–the range of visual acuity values of the cataract-treated participants. Nonetheless, to ensure the procedure would lead to the desired contrast sensitivity function (CSF) for each participant, we tested the visual acuity also of the sighted controls. We visually inspected the CSF of each sighted participant (tested with visual blur) and we included in the study only those whose CSF matched the desired CSFs in terms of both cut off frequency and shape. In other words, when the CSF of a sighted control did not match the one of the to-be-matched cataract-treated participant (in the cut off frequency and/or in the shape of the function), that sighted control was not included in the study. This led to excluding 8 sighted controls, before reaching the final sample of 20 controls, individually matched to the cataract-treated participants. We have now reported these further details in the paragraph entitled ‘Procedure to blur vision in sighted controls’ (Materials and Methods). Moreover, we have provided a Figure in the Supplemental Material, showing the mean CSF in the group of cataract-treated participants and in the group of sighted controls tested with visual blur (Figure supplement 1). In that figure, it is possible to appreciate that we ensured matching the two groups not only for the cutoff frequency, but also for the shape of the whole function. However, we have now also mentioned in the Discussion that we cannot exclude that other possible visual differences, besides spatial visual acuity, that we did not consider, between the group of cataract-treated and that of controls tested with visual blur might have influenced the recalibration performance.
Another more minor or technical issue is some lack of detail in how the calibration index, which feeds into most of the key analyses, is calculated. It is likely that many different ways of doing this would lead to similar conclusions, but it should be clear, including for the sake of replicability.
While the index is briefly mentioned in the Results section, we have now explained it in detail in the Material and Methods section. This recalibration index combined the amount of recalibration in the prism phase and at the beginning of the post-prism phases (Adaptation and Initial Aftereffect, respectively). Adaptation was calculated as the error reduction in the prism phase (the induced prism distortion–11.31°–minus the average of the last three pointing errors of the prism phase, cf. Fortis et al. (2010)). Initial Aftereffect was calculated as the magnitude of the aftereffect exhibited right after prism removal (i.e., average of the first three pointing errors of the post-prism phase). The Initial Aftereffect was correlated with the amount of Adaptation in the prism phase (see Material and Methods) and thus provides converging information which in order to increase power can be summarised in the recalibration index. That is, the recalibration index irecal was calculated as the average between Adaptation and the (negative) Initial Aftereffect. Such index is normalized on the induced prism distortion (i.e., the index is divided by 11.31°), so that it ranges between 0 and 1. Further details are provided in the Material and Methods section.
Reviewer #2 (Public Review):
It is very interesting that recalibration effects in the cataract-reversal group increase over time. However, it seems as if the conclusion that it takes about two years to reach recalibration effects comparable to those of typically sighted controls is based on repeated measurements of two participants tested 2 and 3 years after their surgery as well as on singular measurements of two participants tested 10 years after their surgery. Close inspection of Figure 1F suggests that four participants reached comparable levels in their second testing session already about 6 months after surgery. Consistently, the confidence interval of the time constant b is rather large (it also seems to differ between the main text and the figure caption). Given this high degree of uncertainty around the time estimate it would be advisable to not report and discuss a fixed duration of two years but rather focus on the increase of recalibration effects and report an interval during which recalibration effects might reach asymptotic levels.
We thank Reviewer 2 for drawing our attention to this important point. Following this advice, we have now discussed the high inter-subject variability in the recalibration performance over time, and we have discussed the uncertainty inherent in the estimate of the rate of improvement leading to a performance comparable to healthy controls within about 2 years - this estimate for sure is very uncertain (see Results and Discussion).
It is important to note that the exponential fit on all measurements (Figure 1F, dark green curve) is not driven by the 2 participants tested more than 10 years after surgery: when excluding them from the exponential fit, the time constant b (b=1.5, 95% CI=[0.39, 2.67]) is comparable to the one obtained in the whole sample.
We have also reported the linear correlation between time since surgery and recalibration index in the first testing session without the 2 participants tested more than 10 years after surgery, as they would drive the correlation. Note that the effect of time since surgery is evident even when removing them from this analysis (main text, red line in Figure 1 F, and Material and Methods). Importantly, also the linear fit on the first test session alone (excluding the participants tested more than 10 y after surgery) provides converging evidence of the fact that the performance level of controls (tested with visual blur) is reached at roughly 2 years from surgery, as visible in Figure 1F (red regression line crossing dashed line of controls).
Regarding the time costant b previously reported in the figure caption, this was related to the inlaid reported in Figure 1 F in the last submission (i.e., the exponential fit on the difference between each pair of cataract participants and controls). We have now removed this inlaid from the figure and its relative fit (in the figure and figure caption) to avoid confusion.
Having longitudinal data from several participants is great and can provide interesting insights. However, to get an idea about the role of visuo-motor experience it would be helpful to not collapse across the different time points for the second evaluation in the depiction of the data and their analysis. Moreover, it would be helpful to have an idea of the degree of variability across repeated measurements in control participants.
We decided to report these data in two ways: 1) In agreement with Reviewer 2, we showed these longitudinal data in their different time points (Figure 1F), so that the progression of the recalibration ability over time after surgery would be more transparent and easier to appreciate; 2) We still present these data also collapsed in Figure 1 E, because we believe this representation helps clarity and completeness: given that we also included the pre-surgical assessment in that figure, it is easier to visually appreciate the differences between pre- vs. first post-surgical assessment and second post-surgical assessment in the re-tested participants. We also rearranged the text accordingly. However, if the Reviewer still believes that this way of reporting the results is unclear or redundant, we will remove Figure 1E.
Unfortunately, we were unable to collect comparable repeated measures from the control children with the same temporal gap between the first and second test.Visuo-motor adaptation and aftereffects are related but clearly separate phenomena not least because visual feedback about the position of the finger was only present during the adaptation phase. Combining both effects into one index potentially obscures differential effects of developmental vision on the processes underlying either phenomenon. This concern is supported by the result that the manipulation of visual precision in typically developed controls affected visuo-motor adaptation and aftereffects differentially. Thus, it would be preferable to drop the combined index and analyze adaptation and aftereffects separately throughout. This will have the additional advantage of allowing for direct comparisons of both effects to those reported in the extensive literature on the topic.
We are grateful to Reviewer 2 for bringing this important point to our attention. We have now run all the correlational analyses separately for adaptation (i.e., error reduction in the prism phase) and aftereffect (mean systematic error in the post-prism phase). We have described these analyses in the Results section and in the Material and Methods section. However, as these separate analyses led to comparable results for adaptation and aftereffect, we did not report them in detail in the main text, as they would be very redundant. While it is possible to appreciate each of them in detail in the Supplemental Materials (Figure 1– figure supplement 4), in the main text we avoided this redundancy by combining them into a unified measure, the recalibration index (irecal). Reviewer 2 is right in highlighting the difference between adaptation and aftereffect. Note, however, that the recalibration index does not include the entire aftereffect (which may have a different time constant as it may well be distinct from the adaptation), but only the amplitude of the initial three trials of the aftereffect after removing the prism (i.e., the mean of the first three pointing errors of the post-prism phase). This initial amplitude of the aftereffect (that we have now called “Initial Aftereffect”) is highly correlated with the amount of recalibration in the prism phase. We have now discussed this point in the Results section. In other word, the recalibration index did not include the aftereffect in the entire post-prism phase (i.e., the systematic error across all trials of the post-prism phase). In fact, we agree with the Reviewer that including the development of the after effect across all trials of the post-prism phase would have potentially shown a different phenomenon, namely the effect of proprioception while reinstating the usual sensorimotor mapping. Indeed, at odds with the prism phase, the pointing task in the post-prism phase was performed in the absence of any optical distortion and in the absence of visual feedback. The development of the aftereffect across all trials of the post-prism phase is analysed in the main text and in Figure supplement 3, while the correlations between each factor (age, visual acuity, etc.) and the mean aftereffect across all trials of the post-prism phase is reported in Figure supplement 4. We have now also clarified all these points in the main text and in the Materials and Methods.
The absence of a significant statistical effect does not provide evidence for the absence of the effect. This problem arises in several instances throughout the paper. For example, a non-significant Kruskal-Wallis-Test does not indicate a similar distribution of baseline pointing errors. A figure showing the distribution of pointing errors from this phase provides far more convincing evidence (l. 134). A non-significant t-test does not provide for the absence of a relation between the change in recalibration effects and visual acuity (l. 225). Here, it would be correct to state that there was no statistically significant difference between visual acuity at the two different post-tests.
The problem that the absence of statistical effects does not allow for any conclusions is even more evident for the correlational analyses, which are severely underpowered. The non-significant correlations should be reported in the supplement rather than in a prominent position in the manuscript and all conclusions based on non-significant correlations must be dropped.
We have now modified the text and Figure 1 accordingly, by rephasing the text and removing the non-significant correlations from the figure.
Figures 1C and 1F suggest that the significant correlation between the time since surgery and recalibration effects might be driven by outliers. The analysis should be repeated without outlier data to make sure that the effect is present in the data.
As reported in the first response to Reviewer 2, we have now re-run the analyses also without the participants tested more than 10 years after surgery. The effect of time since surgery is present even when removing the outliers (See main text and Figure 1F).
The abstract makes rather general claims about the influence of developmental vision on recalibration and plasticity which are not supported by the data. All conclusions should be restricted to the visuo-motor domain, which in my view will not impact their importance.
We thank the Reviewer for the comments, and we have adapted the abstract accordingly.
Given that most participants had residual light perception, it would be more accurate to consistently speak of absent pattern vision rather than visual deprivation.
We have rephrased the text accordingly.
-
Evaluation Summary:
This paper will be of interest to researchers in the fields of motor control, visual perception, learning and brain plasticity, sight loss and rehabilitation. The paper shows the contributions of sensory-motor experience to the development of visuo-motor recalibration abilities using careful experimental methods and analyses, comparing a rare population of late-operated cataract patients with control groups.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)
-
Reviewer #1 (Public Review):
The authors asked to what extent early visual and visuomotor experience is essential for developing the ability to recalibrate the visuo-motor system flexibly. This kind of recalibration crucially underpins everyday actions, allowing the brain to issue effective feed-forward motor control commands that correctly account for temporary changes in sensory-motor mappings (e.g. when using tools, carrying objects, wearing new glasses). To address the role of experience in developing these recalibration abilities, they used the unusual clinical population of late-operated cataract patients: children and adolescents who initially had many years of sensory experience that is atypical in that it lacked effective pattern vision. They used a standard sensory-motor task in which participants point to targets with and …
Reviewer #1 (Public Review):
The authors asked to what extent early visual and visuomotor experience is essential for developing the ability to recalibrate the visuo-motor system flexibly. This kind of recalibration crucially underpins everyday actions, allowing the brain to issue effective feed-forward motor control commands that correctly account for temporary changes in sensory-motor mappings (e.g. when using tools, carrying objects, wearing new glasses). To address the role of experience in developing these recalibration abilities, they used the unusual clinical population of late-operated cataract patients: children and adolescents who initially had many years of sensory experience that is atypical in that it lacked effective pattern vision. They used a standard sensory-motor task in which participants point to targets with and without displacement of the visual image via a prism lens: after the prism displacement, the visuo-motor mapping needs to be recalibrated to enable effective pointing. They compared late-operated cataract patients with controls matched in age, controls matched in both age and visual acuity (via added visual blur), as well as an extensive broader comparison group of typically developing 6- to 17-year-olds. Their key findings were that recalibration was less effective - both in the initial effect and in the subsequent after-effect - in the patient group than in control groups; this was not related to chronological age but was related to time post-operation, such that performance came to match controls after around 2 years of improved visual experience. The authors conclude that flexible sensory recalibration abilities normally rely on extensive sensory-motor experience in childhood, and suggest that the underlying computational problem is establishing the correct correspondences between sensory and motor coordinate frames. This may be achieved through extended exposure to the sensory consequences of self-generated movements.
Strengths of the approach include use of the established (although rare and difficult to access) model population of late-operated cataract patients and a well-established experimental task (pointing after displacement of the visual image by viewing through prism lenses). The task has a known typical time-course of behaviour - supplemented here by an extensive additional study on typical development using the exact same main task, which even alone would be a meaningful contribution to literature on sensory-motor development. The procedure, measures, analysis, and the approach to control groups are careful and rigorous. The findings are rich in showing not only an initial deficit in patient vs control groups but also an approximate time course for further learning and development after which point (by ~2 years) the patients come to match controls. A challenge is the heterogenous group, in terms of age at operation and ages at testing and follow-up. However, this is very usual and almost inevitable in the literature with this kind of population, and is dealt with well in the analyses. The approach is also well supplemented by repeated follow-up of a portion (actually more than half) of the group.
One potential issue is the role of baseline pointing precision differences across the groups. It would be useful to better understand the potential role of the reduced pointing precision that was found in the cataract group (Supplemental Figure 1B). It is not surprising that, following visual deprivation, this group's predictive feedforward visuo-motor control was less precise than that of controls, even in the baseline measures before any prism manipulation, and even when the controls' vision is comparably blurred. It seems likely (although is not shown) that during the adaptation phase and the post-adaptation phase, the variability of individuals around their (gradually shifting) mean pointing location would also be higher than in controls. I wonder how large an explanatory role there could be simply for this noisier initial visuo-motor mapping in the patient group. It might be said that, on each trial, they intend to carry out a feedforward plan with a certain endpoint, but because of noise, they are on average substantially further from that endpoint than comparable controls are. So, during recalibration, while controls are dealing mainly with cancelling out one kind of error - the constant error due to the prism adaptation - the cataract patients are also dealing with more variable errors due to their own noisier visuo-motor system. In theory, could this alone - higher initial noise in the system - explain the difference? This seems like a simpler explanation than that the system has developed differently in substantial ways to do with its abilities to learn and adapt. One starting point for checking in to this would be asking if initial pointing variability predicts recalibration (perhaps controlling for visual acuity), both at first test and in the repeated participants. Another would be looking into ways to perturb controls' baseline pointing performance further (perhaps with something like an unexpected added weight rather than more visual blurring) so that their variable pointing errors were matched to the cataract group.
Another question is how well the contrast sensitivity function (CSF) as a whole (not just the maximum acuity point) was matched - this is dealt with only briefly. I am not sure to what extent the blurring manipulation would be expected to change the shape of the CSF as a whole to be in line with that of patients, and to what extent other aspects of the CSF besides the maximum acuity point determine the precision and accuracy of ballistic pointing movements under the experimental and lighting conditions used in the study. Depending on the answers to these questions, the concern could be that visual differences relevant to control of pointing remained across the patient and blurred control groups.
Another more minor or technical issue is some lack of detail in how the calibration index, which feeds into most of the key analyses, is calculated. It is likely that many different ways of doing this would lead to similar conclusions, but it should be clear, including for the sake of replicability.
-
Reviewer #2 (Public Review):
It is very interesting that recalibration effects in the cataract-reversal group increase over time. However, it seems as if the conclusion that it takes about two years to reach recalibration effects comparable to those of typically sighted controls is based on repeated measurements of two participants tested 2 and 3 years after their surgery as well as on singular measurements of two participants tested 10 years after their surgery. Close inspection of Figure 1F suggests that four participants reached comparable levels in their second testing session already about 6 months after surgery. Consistently, the confidence interval of the time constant b is rather large (it also seems to differ between the main text and the figure caption). Given this high degree of uncertainty around the time estimate it would …
Reviewer #2 (Public Review):
It is very interesting that recalibration effects in the cataract-reversal group increase over time. However, it seems as if the conclusion that it takes about two years to reach recalibration effects comparable to those of typically sighted controls is based on repeated measurements of two participants tested 2 and 3 years after their surgery as well as on singular measurements of two participants tested 10 years after their surgery. Close inspection of Figure 1F suggests that four participants reached comparable levels in their second testing session already about 6 months after surgery. Consistently, the confidence interval of the time constant b is rather large (it also seems to differ between the main text and the figure caption). Given this high degree of uncertainty around the time estimate it would be advisable to not report and discuss a fixed duration of two years but rather focus on the increase of recalibration effects and report an interval during which recalibration effects might reach asymptotic levels.
Having longitudinal data from several participants is great and can provide interesting insights. However, to get an idea about the role of visuo-motor experience it would be helpful to not collapse across the different time points for the second evaluation in the depiction of the data and their analysis. Moreover, it would be helpful to have an idea of the degree of variability across repeated measurements in control participants.
Visuo-motor adaptation and aftereffects are related but clearly separate phenomena not least because visual feedback about the position of the finger was only present during the adaptation phase. Combining both effects into one index potentially obscures differential effects of developmental vision on the processes underlying either phenomenon. This concern is supported by the result that the manipulation of visual precision in typically developed controls affected visuo-motor adaptation and aftereffects differentially. Thus, it would be preferable to drop the combined index and analyze adaptation and aftereffects separately throughout. This will have the additional advantage of allowing for direct comparisons of both effects to those reported in the extensive literature on the topic.
The absence of a significant statistical effect does not provide evidence for the absence of the effect. This problem arises in several instances throughout the paper.
- For example, a non-significant Kruskal-Wallis-Test does not indicate a similar distribution of baseline pointing errors. A figure showing the distribution of pointing errors from this phase provides far more convincing evidence (l. 134). A non-significant t-test does not provide for the absence of a relation between the change in recalibration effects and visual acuity (l. 225). Here, it would be correct to state that there was no statistically significant difference between visual acuity at the two different post-tests.
- The problem that the absence of statistical effects does not allow for any conclusions is even more evident for the correlational analyses, which are severely underpowered. The non-significant correlations should be reported in the supplement rather than in a prominent position in the manuscript and all conclusions based on non-significant correlations must be dropped.Figures 1C and 1F suggest that the significant correlation between the time since surgery and recalibration effects might be driven by outliers. The analysis should be repeated without outlier data to make sure that the effect is present in the data.
The abstract makes rather general claims about the influence of developmental vision on recalibration and plasticity which are not supported by the data. All conclusions should be restricted to the visuo-motor domain, which in my view will not impact their importance.
Given that most participants had residual light perception, it would be more accurate to consistently speak of absent pattern vision rather than visual deprivation.
-
