Beacon Hill Study: An Example of Analyzing ResearchA. Boiler Plate for analyzing research (or writing research)1. What is the topic or issue? Why should we care?2. Background and discussion of the issues.a. review of literatureb. development of questions which the research will answer3. Development of Model to Be Estimateda. specification and discussion of model:Dependent variablesVariables of interestControl Variablesb. other issues in specification of model (functional form, etc)4. Data Sources:a. where was the data obtainedb. how were variables constructed and specified (how does thisrelated to the model)c. exclusions and sample size5. Description of the underlying data (pre-regression)6. Presentation of Estimatesa. Tables of estimates (coefficients, standard errors, t-stat against 0)b. r2, r2-bar, clear description of units or other issues to allow readersto understand the material7. Discussion of results and implicationsa. Were estimates consistent with the hypothesis?b. What can be learned from these estimates?8. Conclusions:a. re-iterate issue and fundamental resultsb. implications, limitations and extensions.B. Beacon Hill 2:1. New data: went from XX to YY observationsDid the population change?2. How did the means of the data change3. How did the estimates change?4. Why does this present a problem for us in terms of learning from BeaconHill studies?C. Effect of PLAs on the Cost of New England School Construction1. Issues:SpecificationNew vs. renovationSource of data (dodge vs. actual)Specification of Dependent variableStability of Estimates to reasonable change in sample.2. Method of data collection3.
View Full Document