Statistical analysis

I have two datasets of yamuna river. I want you to analyze those datasets using:
Descriptive Analysis
1. Tabulate the data
2. Find averages if multiple observation for one parameter is available
3. Compare the values by three different time
4. Compare the values by 7 different sample locations
5. Study variance across locations and time (dispersion)
6. Frequency of certain parameter with location
7. Create appropriate infographics (Box plot, bar/pie/column charts etc.)
8. All the above will be performed using Microsoft Excel
Correlation Analysis
1. Appropriate correlation analysis will be conducted across variables (Mostly Pearson
correlation)
o Find the relationship between variables (Excel)
o Scattered plot with fitted line/curve to show correlation (Excel)
2. If require create correlation matrix (R or Python)
3. Comparison testes
o Compare both Physio Chemical and Metal Ions at different time independently –
Seasonal effect (Method is data specific)
o Also compare both Physio Chemical and Metal Ions in different locations
independently (need information which are upper segment and Himalayan segment
& method will be data specific)
4. Check the properties of data set and assumptions of statistical test. (Excel)
o Select appropriate method based on assumption (It’s t-test in reference paper, there
are some assumptions for t-test. Normality, homogeneity of variances etc.)
o If the data fulfills assumptions of t-test then we can use it, other wise adopt nonparametric alternatives (Wilcoxon Signed Ran Test etc.), (Excel/R/python)
5. Valuate and interpret the result from critical value, calculated value and p value etc.
6. Decide whether to reject null hypothesis or not.
Regression Analysis
1. Need information regarding what is the final water quality index.
2. Fix the above as dependent variable.
3. Correlation analysis to select independent variable. Correlate each variable with final water
quality index (Excel)
4. Check assumptions of Regression model.
o If assumptions are satisfied, then perform multiple linear regression analysis.
o If not try to transform the data to meet assumptions (logarithmic, square box-cox
transformation etc.) (Excel/R/Python)
5. If transformation is not possible then adopt non-parametric alternatives, bootstrapping etc.
(R)
6. Interpret the results with infographics. (Excel/R).
Multivariate Statistical Analysis
1. Check assumptions of ANOVA (One-Way). To assess difference among one parameter (either
location or time)
2. If required two-way ANOVA: difference among both location and time studying at a time.
NB: Objective not clear.
3. Cluster analysis or Principal Component analysis: Data specific.
4. It seems both may not be required according to objectives and wildness of dataset.