The swimming pleasure measure (SPM) is an invention aimed at finding out how people in body and mind (as biological sensor and as subjective arbitrator) perceive and experience publically managed recreational waters, and how such a metric might compare with other parameters that are usually collected and evaluated by environmental monitoring systems.

There are three kinds of SPM refered to below: The spm_interview is the voice of the crowd describing the pleasure of being at the beach, spm_water_only is the voice of the crowd describing the pleasure of specifically being in the water. The spm_effective is the application specific 'human computer resource percept', the combination of the subjective interview sourced data and the sensor data as described in the 'numeracy' section. The spm_effective is a new class of environmental experience metric that is not recorded by any environmental monitoring discipline.

In order to see whether the proposed metric is effective we analyzed our SPM and sensor data (comprising over 200 unique entries in 2010, and over 600 unique data entries in 2012) with different statistical tests (see below). The SPM values stem from interviews of swimmers immediately after exiting the water, still wet from the swimming experience. We asked swimmers to map their personal experience onto a scale of 1(worst) - 10 (best) and recorded the results via mobile phone application directly into a remote database . The sensor data stem from state of the art chemical environmental sensors as well as NOAA weather data and a desktop bioincubator capable of detecting e-coli, the main indicator of contaminated beach waters, at the single coliform level. (See data sources and data types under 'numeracy').

Analysis shows surprisingly that no combination of our sensors from water chemistry, water biology and weather correlate significantly with the SPM. While warm water and air were a given in our summer tests, wind, insolation, rain, turbidity and the non-perceivable parameters of chlorophyll, pH and E. coli were not. This suggests that the personal experience of the quality of the water while going for a swim is largely independent of how government agencies report on water quality. Additional analysis shows that the pleasure of being at the beach is, not surprisingly, significantly associated with the pleasurable experience of the water itself. Analysis also shows that there are significant differences between different beach visitor groups. For example, men and visitors (non-locals) are more likely to give a higher SPM score. Teenagers are the most discerning beach visitors. They routinely offered lower SPM scores.

Most interesting however, is evidence of a significant association between perceived water odor (also collected during the interviews) and SPM. This means that a significant conduit into the subjective experience of water quality and the pleasure of being in the water is the water odor. While this might seem intuitive, it is an important result as the usual selection of physical water parameters for beach monitoring and environmental health of beach waters do not include water odor; it is not simply considered. Our data suggest that water odor is an important part of the pleasure of being at the beach and directly impacts people's intuitive perception of the health of a body of water.

The table below summarizes the data analysis of the SPM:

statementtestevidencelocationyearsample size
there exists a positive relationship between spm and water odorKruskal-Wallisp-value < 0.0001
p-value < 0.0001
beaver island2010 2012N=203
N=664
visitors are more likely to give higher spm water scores than local residentsWilcoxon rank sump-value = 0.0064beaver island2010N=32
men tend to give different spm water values than womenWilcoxon rank sump-value = 0.0193beaver island2012N=644
men are more likely to give higher spm water scoresWilcox rank sump-value = 0.0193beaver island2012N=644
different age groups perceive the spm water differentlyKruskal-Wallisp-value = 0.0073beaver island2010N=200
teenagers are more likely to give lower spm water scores than adultsKruskal-Wallis pairwise comparisonsp-value = 0.0008beaver island2010N=200
there is 95% confidence that 28% to 36% of people on the beach are concerned about the cleanness of the beach.one sample normal testbeaver island2012N=580

As others we also wanted to know whether the E. coli data is correlated with environmental sensor data. In our case we found meaningful relationships between E. coli and dissolved oxygen (DO) as well as wind direction. At Beaver Island and Woodlawn Beach, winds from the south (including south-east and south-west) together with elevated DO levels can indicate increased levels of E. coli concentration. The 2008 Woodlawn results are based on only 4 data points, hence statistically not significant. However, they do support the observation that in the Buffalo areas tested, DO and wind direction are important factors to consider as approximators for potential E. coli problems.

The table below summarizes the technical data analysis for E. coli:

statementtestevidencelocationyearsample size
dissolved oxygen and wind direction(south) are significant predictors for E.coli levels.Multiple regression with Cp model selection criteriap-value DO = 0.0386
p-value wind dir = 0.0008
adjusted Rsquared = 0.3417
beaver island, woodlawn beach2008 2012N=4
N=26

The p-value is an estimate of the probability that the result has occurred by statistical accident. A large p-value represents a small level of statistical significance and vice versa. In this case, an increase in spm_water_only score is significantly associated with increase in spm score, for example.

1 2 3 4 5 6 7 8 9 10
team publications presentations contact
sitemap