Test for Heteroskedasticity with the White Test
How to Distinguish between Homoskedastic and Heteroskedastic Disturbances
How to Choose a Forecasting Method in Econometrics

10 Components of a Good Econometrics Research Project

Following are the ten components you need to include in any econometrics research project. No matter what the specifics of your class assignment, you’ll probably be expected to come up with a topic, collect data, use econometrics software to complete the analysis, and interpret your findings.

Introducing your topic and posing the primary question of interest

The first paragraphs of your research paper should provide an interesting description of your topic. This section is important because it either captures your readers’ attention or bores them right from the start.

The introductory section of your research project should include the following two components, in this order:

  • Explanation of the topic

  • Description of your approach

Discussing the relevance and importance of your topic

The introductory section of the paper should also motivate the subject so that readers appreciate the importance of the topic and your findings.

The first paragraph of your introductory section should provide a basic explanation of your research question to spark the reader’s interest, and you should follow it up in the second paragraph with a more profound argument for the importance and relevance of the topic.

Reviewing the existing literature

Other researchers are likely to have examined the topic of your paper (or something closely related), so one section of your paper should review other research on the topic. The length of this section depends on the amount of previous research that’s been completed on your topic, but you should plan on about two to four pages of literature review.

This section should be placed immediately after introducing the topic and briefly describing your contribution in the introduction, but before you begin getting into the details of your model and data.

In your literature-review section, focus on summarizing, highlighting the strengths, and pointing out the weaknesses of prior research. Unless the goal of your work is to replicate or update an existing study with new data, you probably want to focus on one of the weaknesses in the prior literature that you intend your own econometric work to address.

Describing the conceptual or theoretical framework

One of the characteristics that differentiates applied research in econometrics from other applications of statistical analysis is a theoretical structure supporting the empirical work. In other words, the theoretical structure from your knowledge of economics is emphasized in econometrics (and should justify the connection between your dependent and independent variables) rather than focus only on the statistical fit between variables.

By tapping into your vast stores of common sense and using solid economic theory, you can come to methodical conclusions about which variables are independent and can be used to explain your outcome of interest.

Explaining your econometric model

After you develop the theoretical structure of your model, you need to connect that with your empirical approach (that is, your method of statistical analysis and observation), which is formally known as your econometric model.

Economic theory guides your choice of dependent and independent variables. At this point, however, you should explain and justify any specification characteristics of the econometric model (logs, quadratic functions, qualitative dependent variables, and so on) that aren’t directly addressed by the conceptual framework.

Discussing the estimation method(s)

Because estimation usually assumes that certain statistical conditions hold, going from your econometric model to estimation may not be entirely straightforward.

Estimation problems arising from a failure of one (or more) of the classical linear regression model (CLRM) assumptions are common in applied econometric research. If the empirical model has potential problems — such as multicollinearity or heteroskedasticity — you should describe the source, discuss how your results may be affected, and explain how you’ll address the complications.

Providing a detailed description of your data

Your econometric results are only as good the data used to estimate your model(s). Give a thorough description of the data you use. Address these issues:

  • How the dataset was acquired and its source(s)

  • The nature of the data (cross sectional, time series, or panel)

  • The time span covered by the data

  • How and with what frequency the data was collected

  • The number of observations present

  • Whether any observations were thrown out and why

  • Summary statistics (means, standard deviations, and so on) for any variables used in your econometric model(s)

Approximately one paragraph of your research paper should describe the content of the data and convince readers that its use is sensible for your research question. In an additional paragraph or two, use quantitative summary statistics to persuade readers that the data is reliable and of high quality.

Constructing tables and graphs to display your results

Most econometric research projects involve estimating numerous variations of related models. After you choose which results are most important and relevant to addressing your research question, you need to organize them in a concise manner.

A useful table typically contains estimates from several different yet related models. It can help convince readers that your results are robust, or it can lead into a discussion about why they’re sensitive to changes in specification.

Interpreting the reported results

Readers may lose track of details regarding the specification of your econometric model, the scale of the variables, and other aspects that influence how your results should be interpreted.

Reporting your econometric results is not enough; you also need to decipher the results for your readers. The most important element in the discussion of your results is the evaluation of statistical significance and magnitude for the primary variables of interest (the ones most important in addressing the research question).

Some of your variables may be more difficult to understand (because, for example, they’re measured in logs, or the model is nonlinear), so you need to provide an interpretation of the coefficient estimates for your readers.

Summarizing what you learned

The conclusion of your research project should synthesize your results and explain how they’re connected to your primary question.

When you summarize your work, begin by explaining what you did in your analysis. Then discuss what you discovered and the implications of those discoveries. Finally, express some limitations of your research (without being too critical) and make some suggestions for future research on the topic.

Be sure to avoid these common mistakes when drawing your conclusions:

  • Focusing on variables with coefficients that are statistically significant even when the magnitude of their effect on the dependent variable is negligible (nearly no effect)

  • Ignoring variables with statistically insignificant coefficients, particularly when this finding contradicts prior beliefs or expectations

  • Add a Comment
  • Print
  • Share
blog comments powered by Disqus
Defining and Justifying the Least Squares Principle
How to Calculate Parameters and Estimators
3 Main Linear Probability Model (LPM) Problems
Bivariate or Joint Probability Density and Econometrics
The Linear-Log Model in Econometrics

Inside Dummies.com