A Society Survey
UNDERSTANDING SOCIETY SURVEYThe Understanding Society is a longitudinal survey (i.e. re-interviewing the same individuals / households over time). It is also commonly referred to as the UK Household Longitudinal Survey (UKHLS). The survey builds on and extends an earlier survey that was known as the British Household Panel Survey (BHPS). The project will involve using a single cross-section (Wave 10) of the UKHLS data as applied to individuals although some household variables are also included in the datafile.The Understanding Society survey provides information on household composition, employment and skills, income and wealth, education, health and lifestyle, social and political attitudes, well-being, environment and transport, children and families. Given its coverage, Understanding Society and its predecessor, the BHPS, have been and continue to be used in a wide range of applications, see for example:https://www.understandingsociety.ac.uk/For the project you will be using a condensed version of the dataset. The dataset will be available to download from Blackboard, but everyone must register with the UK Data Service before using the dataset. Details on the registration process will be provided during week 1. Please note that the dataset will only be available once the majority of students have registered.ASSESSMENT DETAILSBoth the formative and summative assessment is group-based. Details on this will be provided separately – some initial advice is provided on page 4 of this document.There is an opportunity to prepare a preliminary project proposal. The preliminary proposal will include title (topic area) and a list of variables being considered. This will be submitted as a Microsoft FormA project proposal will be submitted which will count for 20% of the final mark for the module. A template to use will be made available. Word limit 750. Submission is via Blackboard.There is an opportunity to submit a research plan which will be used for to provide details on the choice and construction and empirical analysis that is being considered. This will be submitted as a Microsoft FormAn econometric project will count for 80% of the final mark for this module. The econometric work is to be carried out using STATA. Word limit 4,000 including tables and graphs but excluding the bibliography and appendix. A suggested outline for the project is provided below and sample projects can be found on the Norwich Economic Papers series (however please note that some of these examples use different statistical software).SUGGESTED OUTLINE OF THE FINAL PROJECTIntroductionState clearly the research question being investigated. Provide motivation in order to convince the reader why the question being addressed is important. Briefly indicate how the issue will be approached (outline the structure of your answer).Background (Literature Review)Describe the underlying theory (if applicable) and findings of important related literature. Identify any original contribution the study could make.ModelThe project should be based on one or a number of relationships / propositions suitable for statistical estimation and/or testing. This should be related to relevant theory (where appropriate). In this section the main regression equation(s) / model(s) should be presented.DataCarefully define the variables; include a list of expected signs (given word count constraints the table which defines the variables and includes the expected signs can be included in the appendix) and a table of descriptive (summary) statistics. Some preliminary charts and diagrams should also be included.Empirical AnalysisThe analysis of the results is a key stage of the research. It involves estimating an initial model or models, performing appropriate diagnostic tests, reformulating and re-estimating, conducting tests of hypothesis and, finally, prediction and policy implications. Consider including a comparative table of results (in terms of different estimation methods, different samples and / or variables) – basically whatever you think is appropriate given the research topic/question being considered. Make sure the results are interpreted appropriately.ConclusionSummarise findings and discuss their implications (e.g. are there any policy implications associated with the results?). Discuss the limitations of the work and outline possible scope for further work.BibliographyProperly cite references. This must conform to Harvard Classification.PresentationPlease pay attention to the readability of the project, e.g. the formatting of tables, figures, equations and regression output. I expect the main figures/tables/equations etc. within the main body of the project. Additional figures and tables can be included in an appendix (which will not count towards the word count) but remember the marker will not necessarily look at the material in the appendix, so use it wisely!AppendixA do-file of all the commands used in producing the work associated with the project must be included. The appendix can also be used to include additional material; table that defines the variables with expected signs (see above), additional regressions run or regression output from STATA. Please avoid using the appendix as a “dumping” ground! Preliminary AdviceStart EarlyTry not to luxuriate in the comfort of having a deadline which at this stage is almost 4 months away. Contact other members of your group now and meet regularly.Be RealisticUse bibliographic sources and online databases to obtain relevant articles. Be realistic in the number of references used (textbooks, journal articles, working papers, articles in the FT, Economist, etc). Avoid referencing tabloid newspapers.Make the Best Use of Previous Academic LiteratureUse these references to identify approaches, both in terms of the estimation methods used but more specifically in the choice of variables. This will provide you with an idea of the most appropriate variables (and the formed they have been included) used in previous studies.Effective and EfficientTry to work effectively. It is often good practice to make a list of tasks to be carried out before sitting down at the computer – this will enable you to work more efficiently. Think about how tasks might be delegated across members of your group.Be PreparedBe prepared to spend a long time with the data, both in terms of preparing the dataset and the statistical analysis (unfortunately this is inevitable). Try not to lose the original datafile just in case things go wrong! Also, you may have to compromise on the choice of variables, depending on data availability. Save work regularly (and save to multiple locations)! Consider having a shared folder for your group where you can add and edit material, making best use of the collaborative nature of the assessment associated with this module.