Segment 1402--Columbus and LaGrange

Site Analysis- Dissolved Oxygen

Columbus- CRWN Site # 1402.0250, LCRA Site #1402.0300

These sites are located in the middle of river segment 1402. The LCRA monitors at this site six times a year upstream from the CRWN site. The site is monitored by a high school science class, which results in a lack of data in the summer months.

A time span of approximately four years was used for this analysis:

The following table is the descriptive statistics from each site for dissolved oxygen:

Note that there are about two times more CRWN data values than LCRA data values. Both data sets have mean and median values that are very close to each other and small skewness values. The mean of the CRWN data is slightly higher than the LCRA site, probably due to the lack of summer monitoring data. The LCRA data is less skewed (0.08) than the CRWN data (1.75). The range of the CRWN data is much higher than the LCRA data (11.85 vs. 4.70). Also the means of each data set are different by almost 0.5 mg/l.

The descriptive statistics indicate that the two data sets are normally distributed. The following are the cumulative frequency plots and histograms for the two data sets:

The Cumulative Distribution Curves indicate a linear trend in the data for both sites. The CRWN site would have a better linear fit if the outlier values were eliminated.

The Histograms show a good visual indication of the normality of the distributions.

 

t-test results

The t-test is performed on the two data sets to determine if they are statistically different:

The t value of 0.909 indicate that there is no statistical difference between the two data sets.

Error Statistics- Standard Error of the Mean

The above results indicate that in 95 out of 100 similar measurements the mean would lie in the approximate range around the mean indicated in the last column. The range around the mean for the volunteer data, n=63, is 0.161 or the mean DO with 95% confidence is 7.98 ± 0.08 mg/l. The range around the mean for the professional data, n=26, is 0.214 or the mean DO with 95% confidence is 7.70 ± 0.107 mg/l.

The graphical representation of the above analysis:

 

Site Analysis- Dissolved Oxygen

River Segment 1402- LaGrange- CRWN Site # 1402.0400, LCRA Site #1402.0410

The CRWN site is a level 2 site monitored by a high school science class downstream of the LCRA site. The CRWN site lacked summer data, again because of school not being in session. For the following analysis a time span of approximately two years was used.

CRWN data- 10/19/95 to 10/14/97, n=34.

LCRA data- 10/17/95 to 10/2/97, n=13

The following table is the descriptive statistics from each site for dissolved oxygen:

For both data sets the mean and median are very close to each other. The CRWN data has a much higher range (2.5 times more data), a much higher standard deviation and the data is more skewed. The means of the two data sets differ by 1.0 mg/l. The higher mean dissolved oxygen level for the CRWN data is at least partially due to the fact that the CRWN data set lacks summer data when, generally, the dissolved oxygen level would normally be at its lowest.

The cumulative frequency distribution curves are:

The cumulative distribution curves display good linearity, which indicates strong normality in the data distribution.

The histograms:

The histograms show the normality of the distribution well.

 

t-test:

The t value for these data sets is greater than 2. This is due mainly to the difference in the mean values of 1.06 mg/l, the largest difference of all of the sites analyzed. This indicates that the hypothesis that there is not a significant difference between the data sets is not valid. The CRWN data contained one measurement that was considerably higher than any other measurement, 17.3 mg/l. Eliminating that data point, the calculated t-value is 1.969, or slightly less than 2.0. This illustrates the sensitivity of this analysis to "outlier" data. The table below shows the descriptive statistics of the CRWN site with the high value eliminated:

 

Standard Error of the Mean

Below, is the tabular results of the standard error of the mean evaluation of the two data sets:

The variability around the mean is significantly higher for the volunteer site despite the fact that there is more than twice as much data. There is much more variability in the volunteer measurements which may warrant taking a closer look at characteristics of the measurements, such as the consistency in time of day of the measurements and the time intervals that are missing.

The standard error of the mean for the CRWN data with the high value of 17.3 eliminated is as follows:

The range around the mean has been reduced by 0.115 by eliminating the one high value. This, again illustrates the sensitivity of the analysis to "outlier" data. This also illustrates the need to have a strong quality assurance system in place to prevent erroneous data from skewing the data.

Below is the graphical representation of the standard error of the mean analysis for LaGrange (click on the thumbnail):

Go back to Charlie Kaough's web page
Go back to Dr. D.R. Maidment's web page