Ergonomic Study of Sea Kayakers and their Equipment
by Philip Clegg
Abstract
The aim of this study was to examine wether a correlation exists between anthropometric measurements of experienced sea kayakers and the dimensions of their equipment,
and how this might help inform performer choice in future.
The potential importance of this is linked with the growth that paddlesports is experiencing. To that end data was collected with the use of an on-line questionnaire posted on several forums. Data analysis yielded a number of interesting results. Implications and applications of the findings are discussed in relation to coach practice and paddler performance.
Introduction
Paddlesports in general, and sea kayaking in particular, have experienced a substantial growth in participation over recent time (Jaeger, 2006). With more young people taking part in the sport (Prather, 2003), and the fact that individuals within specific recreation activities are inclined to continue participating in these activities well into adulthood (Freysinger, 1999), would appear paddlesports will continue to thrive.
In any sport, a key consideration is how the dimensions of the equipment used have an impact on performance. As a sea kayaking coach, I have directly observed the affects of boat choice on performance issues such as balance, speed and manoeuvrability; and paddle dimension on speed, power and endurance. Added to this, the high cost of sea kayaking equipment, its longevity and storage issues means choosing the right equipment first time around is important for sea kayakers.
The aim of this study was to test the hypothesis ‘There is a correlation between anthropometric measurements of experienced sea kayakers and the dimensions of their equipment'. This is of interest because the results may help paddlers be better informed on how to select their equipment, specifically boat and paddle.
Review of the literature
There have been very few studies in the area of paddlesports, especially sea kayaking. Of the studies that have been completed sea kayaking has been shown to be a growth sport. A study in a wide variety of outdoor activities showed paddlesports to be experiencing the second highest growth (Prather, 2003). In a separate study of paddlesports sea kayaking showed the most growth (Jaeger, 2006). Sea kayakers on Prince William Sound were studied and researchers found that sea kayaking had grown and they expected it to continue (Twardock & Monz, 2000). Sea kayak purchases in North America eclipsed canoe sales for the 2007 season, and were up 144% from the previous year (Morphet, 2008).
This growth has also been shown to be sustainable. In a study of outdoor recreation activities 27% of participants were found to be in the 16 to 24 year old group indicating continued growth. In the same study there was shown to be an increase in the take up of paddlesports by youth (Prather, 2003). Add to this the fact that participants in specific recreation activities tend to continue in these activities when they mature (Freysinger, 1999) and this suggests that sea kayaking will continue to grow in the future.
Although the popularity of sea kayaking is evident through several data sources, there are relatively few studies examining sea kayakers and none studying their equipment. As such I have set out to seek out answers in relation to paddler ergonomics and test the hypothesis that ‘there is a correlation between anthropometric measurements of experienced sea kayakers and the dimensions of their equipment' in order to inform my coaching practices and inform the wider paddling and coaching communities.
Method
In order to accumulate suitable data that would help examine the hypothesis, a questionnaire was compiled. The questions were constructed to try and obtain meaningful anthropometric data from paddlers of a reasonable experience level that could be tested for relationships with the dimensions of their kayaks. The first step involved a pilot session using 10 respondents to check the validity of each question using a paper based questionnaire. After refinements an on-line version of the questionnaire was then built and put online (Appendix 1). Wherever possible the questions were provided with multiple choices options to maintain the quality of the answers given. All the responses sought were directly or indirectly giving empirical data about the participants and their kayaks.
Although only experienced paddlers were invited to fill in the questionnaire the first question asked their level of experience and gave answer choices of beginner, intermediate and experienced. This was used to screen out beginners who were less likely to have refined their kayak choice over time. In this study the data of intermediate and experienced paddlers were both used as it was thought that the experience level of intermediates would still be sufficient to have made informed choices about their kayaks. Questions two to six collected anthropometric data and where necessary a simple measuring guide was included. These measurements were carefully chosen to provide the most meaningful data without making the questionnaire too time consuming. The last two questions was collecting information about the participant’s equiptment, namely make and model. By asking for the make and model of the equipment several empirical measurements could be obtained from the manufactures without making the questionnaire too taxing for the participants.
With the boat make and model, length; width and volume data was obtained where available from the manufactures. In addition to this a category of length multiplied by width was added to the results. This was used to approximately represent hull area which is a statistic none of the manufacturers produce. With the paddle make and model, blade area data was obtained where available from the manufacturers. In addition paddle length statistics were collected from the participants.
Participants for this study were sought from web based sea kayak forums all over the world and the questionnaire was kept online for a period of three months.
Data Analysis
The data was examined for outliers and tested for normality. Spearman’s Rank Correlation Coefficients were carried out across all the categories where relationships were of interest. Correlations that provided p values of 0.01 were considered significant and p values of 0.001 were considered strong correlations. The correlations were displayed in a matrix.
Results
The questionnaire responses were used to create the data fields presented here, from which data were grouped and means and standard deviations calculated. The following table describes the distribution of data within the data fields for the questionnaires returned (Table 1).
|
Category |
Abbreviation |
N |
M |
Standard Deviation |
|
Height |
Hei |
101 |
177.44 |
9.35 |
|
Sitting Height |
Sit |
101 |
89.37 |
6.11 |
|
Arm Span |
Arm |
101 |
179.67 |
11.48 |
|
Shoulder Width |
Sho |
101 |
47.14 |
6.23 |
|
Weight |
Wei |
101 |
87.13 |
67.27 |
|
Boat Length |
B Le |
98 |
518.21 |
25.95 |
|
Boat Width |
B Wi |
98 |
54.12 |
2.09 |
|
Boat Volume |
B Vo |
61 |
317.5 |
38.70 |
|
Boat Length x Width |
BLW |
98 |
28034 |
1581.7 |
|
Blade Area Euro |
BAE |
76 |
666.5 |
44.76 |
|
Paddle Length Euro |
PLE |
82 |
212.5 |
5.20 |
|
Paddle Length Greenland |
PLG |
19 |
219.9 |
7.65 |
Table 1. Numbers, Means and Standard Deviation for the parametric data
Data from 101 participants was collected. This was tested for outliers. 3 participants gave their experience level as beginners. This data was only used to examine the relationship between the experience level of paddlers and the dimensions of their kayaks and not used to examine the relationship between anthropometric measurements of paddlers and the dimensions of their kayaks. 3 participants had boat models for which length and width statistics could not be obtained. 40 had boat models for which volumes could not be obtained. Of the 82 participants that had euro paddles 6 had paddles for which blade area could not be obtained. 19 participants had Greenland paddles.
The data was tested for normality and as only the data for height and arm span were normally distributed Spearman’s Rank Correlation Coefficients were carried out across all the applicable categories and displayed in a matrix in Table 2.
|
|
Exp |
Hei |
Sit |
Arm |
Sho |
Wei |
B Le |
B Wi |
B Vo |
BLW |
PBA |
PLE |
|
Hei |
0.05 |
|
|
|
|
|
|
|
|
|
|
|
|
Sit |
>0.05 |
0.001 |
|
|
|
|
|
|
|
|
|
|
|
Arm |
>0.05 |
0.001 |
0.001 |
|
|
|
|
|
|
|
|
|
|
Sho |
>0.05 |
0.001 |
0.001 |
0.001 |
|
|
|
|
|
|
|
|
|
Wei |
0.05 |
0.001 |
0.001 |
0.001 |
0.001 |
|
|
|
|
|
|
|
|
BLe |
>0.05 |
0.05 |
>0.05 |
0.05 |
*0.01 |
**0.001 |
|
|
|
|
|
|
|
BWi |
**0.001 |
>0.05 |
>0.05 |
>0.05 |
*0.01 |
0.05 |
0.05 |
|
|
|
|
|
|
BVo |
>0.05 |
**0.001 |
**0.001 |
>0.05 |
**0.001 |
**0.001 |
0.001 |
>0.05 |
|
|
|
|
|
BLW |
>0.05 |
*0.01 |
>0.05 |
0.05 |
**0.001 |
**0.001 |
0.001 |
0.001 |
0.001 |
|
|
|
|
BAE |
0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
0.05 |
|
|
|
PLE |
>0.05 |
0.05 |
>0.05 |
*0.01 |
*0.01 |
0.05 |
*0.01 |
0.05 |
0.05 |
**0.001 |
>0.05 |
|
|
PLG |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
>0.05 |
Null |
Null |
Table 2. The Spearman’s Rank Correlation Coefficients p value results.
(Correlations within the areas of interest of this study, a p value of 0.05 was considered a weak correlation, *a p value of 0.01 was considered significant and ** a p value of 0.001 was considered strong. The full results are shown in Appendix 2.)
Discussion
Results from this study present an initial understanding of the correlations between anthropometric measurements of experienced paddlers and the dimensions of their equipment. In terms of the proposed hypotheses for this study, it was proved to be true in many of the correlations and false in others. The interest of this study lies in where relationships were, and to some extent, were not found; and what the relationships reveal.
Boat volume provided the largest number of strong correlations (p=0.001) in the relationships between anthropometric measurements and boat dimensions. This was probably down to the simple fact that larger people need larger boats.
The second highest number of strong correlations between boat dimensions and anthropometric measurements was boat length multiplied by width. It was strongly correlated with weight and shoulder width and also significantly (p=0.01) correlated with height. The weight correlation is perhaps the most interesting as length multiplied by width, although not directly, is closely linked to the area of the hull and it is probably here that the real correlation lies. The area of the hull directly affects the displacement which in turn affects how deeply the boat will sit in the water with a given load. If the boat sits too high or too low in the water the performance will be affected. Currently no manufactures produce hull area statistics yet this could be the most significant anthropometric measurement factor in choosing a boat.
Boat length was strongly correlated with weight, and also significantly correlated with shoulder width. The correlation with weight is probably connected to displacement, as already mentioned. It could be that shoulder width is correlated with boat length independently or that the strong correlation between shoulder width and weight caused the relationship. This study cannot show the cause of this relationship. If the relationship has an element of independence it could be hypothesised that paddlers with wider shoulders are physically advantaged to manoeuvre and propel longer boats.
The strongest relationship with boat width was level of experience. You could predict that as paddlers become more experienced they will be more capable of paddling narrower, less stable boats, however the results don’t show this. It could be that fit and able experienced paddlers make the informed decision that wider boats improve performance or equally that with the increased age that inevitably comes with experience also comes performance reducing factors that make a wider boat more suitable. This second factor is supported by the evidence that, although not significant, there is a weak correlation (p=0.05) with experience and weight, telling us that more experienced paddlers weigh more. Boat width was also significantly correlated to shoulder width. A possible reason is that narrower shouldered paddlers forward stroke is impeded by paddling wider boats. Another possible reason is that because of the stability issues of being ‘top heavy’, wider shouldered paddlers need wider boats.
The paddle blade area statistics could only be collected for euro paddle manufactures as only they provided them. The data did not produce any significant statistics. This could be because there are no correlations, or because the correlations are too complex for the statistical analysis carried out. I would hypothesise that there is a negative relationship between paddle blade area and paddle length as these factors affect the ‘leverage’ of the paddle. I would also suggest that there is a positive relationship between the ‘leverage’ of the paddle and anthropometric measurements of paddlers. In the statistical analysis carried out these two factors could well oppose each other cancelling out any correlations.
The euro paddle length data provided several correlations across both anthropometric measurements and boat dimensions. There were significant correlation between arm span and shoulder width and interestingly there was only a weak correlation with height. Boat length had a significant correlation while boat length multiplied by width had a strong correlation. Traditionally paddler height has been most commonly used when determining appropriate paddle length but this data suggests that arm span and shoulder width could be more appropriate anthropometric measurements. Arm length is strongly correlated to height but I would assume that if any height was a factor in paddle length it would be sitting height as I can’t see that leg length would affect paddle length choice. However there was no correlation found between sitting height and paddle choice. This would suggest that using height to determine paddle length is only taking advantage of the correlation with arm span and so using arm span would be more accurate. While arm span is related to shoulder width because shoulder width is part of the arm span I think the strong correlation with shoulder width could have some independence of its own. Shoulder width is often used to determine the appropriate distance apart to hold the hands on the paddle shaft. In my experience a distance approximately equal to or slightly greater than the paddlers shoulder width is most commonly used. This could explain the strong correlation.
The stronger relationship with euro paddle length was with boat dimensions. The significant boat length relationship was probably due to the fact that longer boats, because they’re faster, require a longer paddle length to maintain the same stroke rate. The strong relationship with boat length multiplied by width may be due to the fact that as well as this, a wider boat also requires a wider, or longer, paddle. This is supported by the weak correlation with boat width. The correlation with boat length and boat width lead to the strong correlation between boat length multiplied by width.
Surprisingly the Greenland paddle length data had no correlations with the anthropometric measurements. This was strange when you consider that the Inuit would specifically use anthropometric measurements when determining their paddle dimensions. Interestingly the Inuit would also use anthropometric measurements when determining their boat dimensions which would undoubtedly lead to a strong relationship between the paddle and boat dimensions which this study suggests is a more important relationship. However the Greenland paddlers in this study didn’t show that relationship either. It is possible that the Inuit methods of determining paddle dimensions don’t apply to western paddlers and their boats. Historically they are a stockier, shorter limbed people due to the evolutionary selection forces of living in a cold climate (Allen 1877) and as such applying their anthropometric measurements may be inappropriate for other people. It is also possibly that the lack of any correlations is due to the fact that greenland paddles are a relatively new introduction to modern sea kayaking and as such there isn’t a culture of education and information when determining the correct paddle dimensions. The small number of participants in the Greenland catergory (N=19) may also be a cause.
Implications for Practice, Limitations and Future Research
The study only looked at the equipment that experienced paddlers used and not what they should use. The statistical analysis carried out showed the relationships or the lack of them but it was not able to explain them. However it still revealed some interesting implications for practice and future research.
I think the strong relationship between paddler weight and boat length multiplied by width could be developed to help to determine boat choice. With more research manufacturers could use a hull area statistic to advise boat choice according to paddler weight. This one statistic would give paddlers a choice of a variety of boats with different lengths and widths from different manufacturers with the same approximate statistic.
The strong correlations between euro paddle length and boat length multiplied by width could be used to re look at how manufactures, coaches and paddlers decide on the appropriate paddle length. This study suggests that arm span and shoulder width are more important anthropometric measurements that the more commonly used height measurement, but more importantly that the boat dimensions, specifically length multiplied by width are the most important. This again supports the need for hull area statistics and more research into the relationship.
This study was limited in that most of the data collected did not show normal distribution and so regression analysis could not be used. I think a study of a larger population would achieve normal distribution and so would be able to use regression analysis to closer examine the relationships. I would suggest a study of at least 200 participants should achieve this. Regression analysis could directly show the relationship
The lack of any correlation with the Greenland paddle length data shows that more information is needed in this area. A closer look at the Inuit practices in determining paddle and boat dimensions, their applications on western paddlers and the decisions made by western paddlers when choosing a Greenland paddle could reveal more information.
Conclusions
This study has revealed an initial understanding of correlations between anthropometric measurements of experienced paddlers and the dimensions of their equipment that as a coach will make me reflect on some of the equipment advice I make for my students.
The relationship between boat length and paddler weight suggests that giving heavier paddlers wider boats, which is often perceived to be the answer, may not be the only solution, increasing the boat length is also important. Equally for lighter paddlers shorter boat choices may be as appropriate as choosing narrower boats. The significant factor is increasing or decreasing the length multiplied by width ratio according to the paddlers weight.
The most significant relationship with euro paddle length being boat length multiplied by width has caused me to completely rethink how I look at paddle choice. Where as in the past I have always based the paddle choice ‘by eye’ on an overall impression of the paddlers anthropometric measurements I will now take into account the boat dimensions in the choice.
However the strongest affect this study has had on me as a coach is to continue the research, specifically in the areas of boat length multiplied by width related to weight and euro paddle length to the point where the relationships can be revealed directly.
Acknowledgements
Thanks to Dr Mark Tozer for his input into the experimental work and assistance while writing it up. Thanks to Annette Burden for her guidance through the analysis and proof reading. Most of all thanks to all the participants who input their data without which none of this research would have been possible.
Appendix
Allen, J. A. (1877). The Influence of Physical Conditions in the Genesis of Species. Radical Review 1, 108-140.
Freysinger, V. J. (1999). Life span and life course perspectives on leisure. In E. Jackson, & T. L. Burton (Eds.), Leisure studies: Prospects for the twenty-first century (pp. 253–270). State College, PA: Venture.
Jaeger, S. (2006). Recreational kayaks dominating sales. Paddlesports Business, Summer, 20–21.
Morphet, S. (2008, August 23). Where to paddle like the pros. The Globe and Mail, p. T2.
Prather, R. (2003, March/April). Paddlesports still growing, oia [outdoor industry association] reports. Paddlesports Business, 1, 6.
Twardock, P., & Monz, C. (2000). Recreational kayak visitor use, distribution, and financial value of beaches in western Prince William Sound, Alaska between 1987 and 1998. In D. Cole, S. McCool, W. Borrie, & J. O’Loughlin (Eds.), Wilderness science in a time of change conference— Volume 4: Wilderness visitors, experiences, and visitor management (pp. 175–180). Missoula, MT: Department of Agriculture, US Forest Service, Rocky Mountain Research Station.
Appendix 1: The content of the questionnaire.
|
1) What is your level of experience in Sea Kayaking? 1 = Beginner, 2 = Intermediate, 3 = Experienced 2) Your height in cm. 3) Your sitting height in cm. (Sitting on the floor with your back against the wall.) 4) Your arm span in cm. (Arms stretched wide, from finger tip to finger tip.) 5) Your shoulder width in cm. (Measure to the ends of the bones.) 6) Your weight in kg. 7) What Sea Kayak do you use for all round use? a) Make b) Model 8) What paddle set up do you use for all round use? a) Make b) Model c) Length in cm |
Appendix 2: The full results.
|
|
Exp |
Hei |
Sit |
Arm |
Sho |
Wei |
|
Hei |
rs101 = 0.213, p < 0.05 |
|
|
|
|
|
|
Sit |
rs101 = 0.157, p > 0.05 |
rs98 = 0.611, p < 0.001 |
|
|
|
|
|
Arm |
rs101 = 0.125, p > 0.05 |
rs98 = 0.807, p < 0.001 |
rs98 = 0.500, p < 0.001 |
|
|
|
|
Sho |
rs101 = 0.153, p > 0.05 |
rs98 = 0.481, p < 0.001 |
rs98 = 0.414, p < 0.001 |
rs98 = 0.38, p < 0.001 |
|
|
|
Wei |
rs101 = 0.225, p < 0.05 |
rs98 = 0.617, p < 0.001 |
rs98 = 0.400, p < 0.001 |
rs98 = 0.469, p < 0.001 |
rs98 = 0.553, p < 0.001 |
|
|
B Le |
rs98 = -0.028, p > 0.05 |
rs95 = 0.224, p < 0.05 |
rs95 = -0.028, p > 0.05 |
rs95 = 0.229, p < 0.05 |
rs95 = 0.273, p < 0.01 |
rs95 = 0.331, p < 0.001 |
|
B Wi |
rs98 = 0.430, p < 0.001 |
rs95 = 0.110, p > 0.05 |
rs95 = 0.090, p > 0.05 |
rs95 = 0.128, p > 0.05 |
rs95 = 0.271, p < 0.001 |
rs95 = 0.208, p < 0.05 |
|
B Vo |
rs61 = -0.194, p > 0.05 |
rs58 = 0.470, p < 0.001 |
rs58 = 0.470, p < 0.001 |
rs58 = 0.161, p > 0.05 |
rs58 = 0.475, p < 0.001 |
rs58 = 0.468, p < 0.001 |
|
BLW |
rs98 = 0.017, p > 0.05 |
rs95 = 0.281, p < 0.001 |
rs95 = 0.045, p > 0.05 |
rs95 = 0.205, p > 0.05 |
rs95 = 0.383, p > 0.05 |
rs95 = 0.481, p < 0.001 |
|
P Ar |
rs76 = 0.246, p < 0.05 |
rs73 = 0.132, p > 0.05 |
rs73 = 0.132, p > 0.05 |
rs73 = 0.065, p > 0.05 |
rs73 = 0.096, p > 0.05 |
rs73 = 0.048, p > 0.05 |
|
PLE |
rs82 = -0.025, p > 0.05 |
rs79 = 0.284, p < 0.05 |
rs79 = -0.003, p > 0.05 |
rs79 = 0.340, p < 0.01 |
rs79 = 0.350, p < 0.01 |
rs79 = 0.262, p < 0.01 |
|
PLG |
rs19 = 0.192, p > 0.05 |
rs19 = 0.284, p > 0.05 |
rs19 = 0.107, p > 0.05 |
rs19 = 0.352, p > 0.05 |
rs19 = 0.337, p > 0.05 |
rs19 = 0.182, p > 0.05 |
|
|
|
|
|
|
|
|
|
|
B Le |
B Wi |
B Vo |
BLW |
PBA |
PLE |
|
Hei |
|
|
|
|
|
|
|
Sit |
|
|
|
|
|
|
|
Arm |
|
|
|
|
|
|
|
Sho |
|
|
|
|
|
|
|
Wei |
|
|
|
|
|
|
|
B Le |
|
|
|
|
|
|
|
B Wi |
rs95 = -0.221, p < 0.05 |
|
|
|
|
|
|
B Vo |
rs58 = 0.630, p < 0.001 |
rs58 = 0.227, p > 0.05 |
|
|
|
|
|
BLW |
rs95 = 0.798, p < 0.001 |
rs95 = 0.340, p < 0.001 |
rs58 = 0.681, p < 0.001 |
|
|
|
|
P Ar |
rs73 = 0.180, p > 0.05 |
rs73 = 0.199, p > 0.05 |
rs73 = -0.024, p > 0.05 |
rs73 = 0.291, p < 0.05 |
|
|
|
PLE |
rs79 = 0.357, p < 0.01 |
rs79 = 0.249, p < 0.05 |
rs52 = 0.332, p < 0.001 |
rs79 = 0.452, p < 0.001 |
rs73 = 0.109, p > 0.05 |
|
|
PLG |
rs19 = 0.215, p > 0.05 |
rs19 = 0.056, p > 0.05 |
rs7 = -0.063, p > 0.05 |
rs16 = 0.315, p > 0.05 |
Null |
Null |