PM608 - Statistical Analysis Water Quality Assignment

Assignment Task

Scenario

There has been a pilot study of the water quality in the water reservoirs supplying the local area. In total, 89 reservoirs supply the area with fresh drinking water. In recent times, the area’s needs have changed significantly in terms of water volumes and quality.

The pilot study aims to understand the key factors affecting water cost and quality. One possible outcome of the study is a programme of re-building the reservoirs to aid water quality. However, this outcome is only likely if the results are significant.

The overall quality of the water is measured using the ‘Index of Biotic Integrity’ or IBI. The highest quality of water will have an IBI score of 100, whereas a score of 0 represents the lowest water quality. Some reservoirs are also ‘treatment sites’ where the water is treated using chemicals to remove impurities before storage in the reservoir. The other ‘non-treatment’ reservoirs use chemicals after the water has left the reservoir.

The data collected for the study is shown below for all 89 reservoirs. A ‘T’ next to the data indicates that the sample was taken from a ‘treatment site’. The water samples were all taken at a standard mean depth of 3 meters however the data collector reported that a small batch of measurements were taken before a fault was detected (and resolved) in the depth gauge. The fault led to some readings being taken at a mean depth of 1 meter. Your analysis will need to take account of any possible anomalies and state their possible impact on any conclusions.

Your task is to analyse the study data and write your recommendations for the Director of Water Supplies in a short, non-technical report (approx. 600 words only). Your results should be clear enough to allow the director to make the best decision. Some examples of possible decisions the director might make are:

• Ask for a more detailed study
• Start new construction
• Change water treatment policies
• Do nothing

Whatever your recommendations are – they must be clearly supported by your calculations using the supplied study data.

You should include a full numerical summary of the data in the study. You should include an explanation of the factors, treatments and any lurking variables which may be present in the pilot study and any future research. You should highlight any possible errors in measurement, data anomalies or outliers and describe their effect on your conclusions. You should describe any correlation between the variables, if present.

You should summarize any findings in the form of clear, supported recommendations that allow the Director to decide policy for the future.

Assessment Approach

Any calculations must only use the methods and techniques described to you on this course.

You must produce a maximum of 2 graphical representations of the given data from the following options:

1. Histogram

2. Box Plot

3. Modified Box Plot

4. Cumulative Relative Frequency

[For example, you may choose: 1 Box Plot and 1 CRF - or - 1 Modified Box Plot and 1 Histogram but no more than 2 in total]

You must also conduct the following two tests (with the given limits in brackets).

1. Linear Regression (with a maximum of 2 correlations, 2 scatter plots and 2 residual plots)

2. 2 Sample Confidence Interval (with a maximum of 2 [2 sample] intervals)

a. Note: depending on circumstance or purpose, there is no ‘best’ confidence level, your choice of confidence level must be justified based on your purpose.

• You should comment on the validity and meaning of any test you conduct in the context of the test result.
• All calculations and statistical tests must be conducted using Microsoft Excel. Use of any other software is not permitted.
• After submission, any spreadsheets you use (your .xls or .xlsx files) must be available to your.

This PM608 - Statistics has been solved by our PhD Experts at My Uni Paper.