Scribd
Upload
18 views

Spatial Data Analytics_coursework

This assignment involves analyzing spatial associations among geographical data by collecting relevant data and conducting regression analyses. Students are required to report their findings in a document not exceeding 2,000 words, including maps, calculations, and interpretations. Key tasks include hypothesis formulation, OLS regression analysis, spatial autocorrelation testing, and potential model re-specification based on results.

Uploaded by

habibur25253
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
18 views

Spatial Data Analytics_coursework

This assignment involves analyzing spatial associations among geographical data by collecting relevant data and conducting regression analyses. Students are required to report their findings in a document not exceeding 2,000 words, including maps, calculations, and interpretations. Key tasks include hypothesis formulation, OLS regression analysis, spatial autocorrelation testing, and potential model re-specification based on results.

Uploaded by

habibur25253
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

SPATIAL DATA ANALYTICS

ASSIGNMENT - Analysing Spatial Association of Geographical Data

Completed pieces of work must be submitted/uploaded as report in the form of a single MS-WORD and
a PDF document with its file size not exceeding 20MB.

The report should include your maps, calculations and results from your analysis, as well as discussion
and interpretation of your results. The report should be no more than 2,000 words long.

The word count for this assignment primarily refers to the length of the main body of the text and
excludes references, table/figure captions, and any numerical information in the tables. Equations in the
main text will count as one word each. Please ensure that figure captions provide information about the
respective figure only and that they do not contain additional information that should be explained in the
main text.

----------------------------------

BRIEFING

This assignment gives you an opportunity to consolidate your knowledge and skills to investigate spatial
association among geographical data. The objective of this type of analysis is to understand
the mechanism of how spatial phenomena/events occur and predict what the outcome will be under
certain scenarios. In this assignment, you will be asked to collect data of geographical events or
phenomena whose attributes are assumed to have causal relationship with each other. You will then
describe your assumption/hypothesis about these possible associations, and carry out non-spatial and/or
spatial regression analyses. Finally, you are expected to interpret the findings about such
associations, and report them in your report of up to 2,000 words.

TASKS

1. Collect geographical data in which you hypothesise the presence of causal relationship between the
attribute variables attached to the geographical features. In your report,
• briefly describe the data you collected;
• specify the dependent and the independent variables to run the regression analysis, and
describe your hypothesis about the causal relationship between these variables.

1
• Present the distribution of the dependent and the independent variables using maps and/or basic
statistics and provide brief comments on possible relationship between these variables (perhaps
through visual observation of the maps).

2. Using GeoDa (or ArcGIS or R), carry out the ordinary least square (OLS) regression analysis (i.e. a
standard, non-spatial regression analysis). Interpret and discuss the results (e.g. is there any
statistically significant relationship between the dependent and the independent variables? Is the
produced regression model a good model, and why do you think that is the case?).

3. Check the presence/absence of spatial autocorrelation in the regression residuals using a global
spatial autocorrelation method. Using GeoDa (or ArcGIS or R), map the spatial distribution of the
residuals, and then carry out a statistical test for spatial autocorrelation on the regression residuals
using Moran’s I index. Interpret the outcome (e.g. does the result indicate the presence of
significant spatial autocorrelation? Does the model require further re-specification and why?).

4. Based on the diagnosis/assessment performed in Step 3, re-specify the regression model as


necessary. You can achieve this either by following one or both of the two solutions shown in the
practical exercise material for the spatial regression analysis topic:
Solution 1 - Find additional independent variable(s) that will explain more of the variation in the
dependent variable and will also eliminate any residual spatial autocorrelation, and
Solution 2 – Find the best fitting “spatial” regression model (e.g. spatial error model) that allows the
residuals to be spatially autocorrelated.
You can apply either one or both of these methods/solutions. Interpret and discuss the results (e.g.
compare the results from the different models you produced and discuss which model you think best
fits the data. How different are the regression residuals between the different models? What is your
conclusion about the relationship between the dependent and the independent variables? ).

ASSESSMENT GUIDANCE

Good marks will be awarded to submissions that successfully demonstrate the following:

1. The data set collected is appropriate for this type of analysis, and a clear and succinct description
is provided for the distribution of the variables used in the analysis.
2. The hypothesis on the relationship between the dependent and the independent variables is
reasonable.
3. The OLS regression analysis is correctly carried out, and the result of the analysis is properly
assessed.
4. The statistical test for spatial autocorrelation in the regression residuals is correctly carried out.
5. The re-specification of the regression model and comparison of the results from the different
models are carried out in a statistically correct manner.
6. Overall results and findings on the relationship between the dependent and the independent
variables are clearly and logically presented.

You might also like