Hypothesis Testing: Economist's Methods [US Focus]

Economists in the United States commonly employ various methods to rigorously test their hypotheses, often drawing upon techniques developed and refined at institutions such as the National Bureau of Economic Research (NBER). Regression analysis, a fundamental tool, allows economists to quantify the relationship between variables, thereby evaluating the validity of a proposed hypothesis within a statistical framework. Furthermore, econometric modeling, frequently implemented using software packages like Stata, provides a structured approach to analyzing economic data and drawing inferences about the relationships between different economic factors. Determining what methods may an economist use to test a hypothesis frequently involves consulting the influential work of figures such as Milton Friedman, whose contributions to positive economics emphasize empirical verification and the importance of testing theoretical predictions against real-world data.

Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic theories. By employing sophisticated statistical techniques, econometrics provides a framework for testing hypotheses, estimating relationships between economic variables, and forecasting future economic outcomes.

This section serves as a foundational overview of econometric analysis. It aims to provide a structured guide to the essential concepts and methodologies that underpin this discipline. We will explore the core principles that enable economists and other researchers to draw meaningful conclusions from complex economic data.

Purpose and Scope of This Overview

This editorial will serve as a focused introduction to the vast field of econometrics, highlighting key areas essential for understanding its application. It is intended as a navigational tool, allowing readers to grasp the fundamental building blocks upon which more advanced econometric techniques are built. The intent is not to provide an exhaustive treatment of every topic. Instead, the goal is to equip the reader with a solid foundation upon which to build further expertise.

The scope will be confined to elements such as hypothesis testing, regression analysis, quasi-experimental methods, and common econometric challenges. This section will not delve into highly specialized or advanced techniques. The focus will remain on the core principles that are most relevant for empirical analysis in economics and related fields.

The Importance of Econometric Analysis

Econometric analysis plays a vital role in informed decision-making across a range of fields. In the realm of public policy, it provides the evidence base for evaluating the effectiveness of government programs. Econometric models can forecast the impact of policy changes on key economic indicators, enabling policymakers to make data-driven decisions.

For businesses, econometrics offers insights into market trends, consumer behavior, and the factors driving profitability. It allows firms to optimize their operations, develop effective marketing strategies, and assess the risks associated with different investment decisions.

Furthermore, econometric techniques are indispensable for academic research in economics. They provide the tools needed to test and refine economic theories, contributing to our understanding of how the economy functions.

In summary, econometric analysis is an indispensable tool for anyone seeking to understand and influence the economic world. Its application spans academic inquiry, policy formulation, and business strategy, making it a critical skill for those seeking to engage with data-driven decision-making.

Hypothesis Testing: The Foundation of Econometric Inference

Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic theories. By employing sophisticated statistical techniques, econometrics provides a framework for testing hypotheses, estimating relationships, and making predictions about economic phenomena. Among these techniques, hypothesis testing serves as a cornerstone, providing the mechanism to evaluate the validity of theoretical claims against empirical evidence.

The Core of Statistical Validation

Hypothesis testing is a systematic process for deciding whether the evidence at hand sufficiently supports a particular belief, or hypothesis, about a population parameter.

It forms the backbone of econometric inference, enabling researchers to draw conclusions about the real world based on sample data.

The process involves formulating two competing statements, the null hypothesis and the alternative hypothesis, and then using statistical tests to determine which one is best supported by the data.

Null and Alternative Hypotheses: Framing the Question

The null hypothesis (H0) represents the status quo or a statement of no effect. It is a precise statement about a population parameter that we aim to disprove.

For example, the null hypothesis might be that there is no relationship between education level and income.

The alternative hypothesis (H1), on the other hand, represents the researcher's belief or the effect they are trying to find evidence for.

It contradicts the null hypothesis and suggests that there is a relationship or an effect.

In the same example, the alternative hypothesis would be that there is a relationship between education level and income. The alternative can be directional (e.g., higher education leads to higher income) or non-directional (there is a relationship, but its direction is unspecified).

The P-value: Weighing the Evidence

The P-value is a crucial concept in hypothesis testing.

It represents the probability of observing data as extreme as, or more extreme than, the data actually observed, assuming that the null hypothesis is true.

In simpler terms, it tells us how likely it is to see the observed results if the null hypothesis is correct.

A small P-value (typically less than a predetermined significance level) suggests that the observed data are unlikely to have occurred under the null hypothesis, thus providing evidence against it.

Conversely, a large P-value suggests that the data are consistent with the null hypothesis, and we do not have sufficient evidence to reject it.

Significance Level (α): Setting the Threshold for Rejection

The significance level (α), also known as the alpha level, is a pre-specified threshold that researchers use to determine whether to reject the null hypothesis.

It represents the probability of rejecting the null hypothesis when it is actually true.

Commonly used values for α are 0.05 (5%) and 0.01 (1%).

If the P-value is less than α, we reject the null hypothesis. This implies that the observed data provide strong enough evidence to conclude that the null hypothesis is false. If the P-value is greater than α, we fail to reject the null hypothesis.

It's important to emphasize that failing to reject the null hypothesis does not mean that it is true; it simply means that we do not have enough evidence to reject it based on the available data.

Type I and Type II Errors: Understanding the Risks

In hypothesis testing, there are two types of errors that can occur: Type I and Type II errors.

Type I Error: The False Positive

A Type I error occurs when we reject the null hypothesis when it is actually true.

This is also known as a false positive.

The probability of committing a Type I error is equal to the significance level (α).

For instance, imagine a researcher tests the hypothesis that a new drug has no effect (null hypothesis) on a disease.

If they reject this null hypothesis and conclude the drug is effective, but in reality, it is not, they have committed a Type I error.

Type II Error: The False Negative

A Type II error occurs when we fail to reject the null hypothesis when it is actually false.

This is also known as a false negative.

The probability of committing a Type II error is denoted by β.

In the same drug example, a Type II error would occur if the researcher fails to reject the null hypothesis and concludes the drug is ineffective, when in reality, it does have a positive effect.

Minimizing Errors: A Balancing Act

Minimizing both Type I and Type II errors is crucial, but it's important to recognize that there is often a trade-off between them.

Decreasing the probability of a Type I error (by lowering α) will increase the probability of a Type II error (β), and vice versa.

Researchers must carefully consider the consequences of each type of error when choosing the appropriate significance level and sample size.

Larger sample sizes generally lead to lower probabilities of both Type I and Type II errors.

The Power of a Test: Detecting True Effects

The power of a test is the probability of correctly rejecting the null hypothesis when it is false.

It is defined as 1 - β, where β is the probability of a Type II error.

A test with high power is more likely to detect a true effect if one exists. Factors that influence the power of a test include the sample size, the significance level (α), and the magnitude of the true effect.

Researchers should strive to design studies with sufficient power to detect meaningful effects. Power analysis can be used to determine the sample size needed to achieve a desired level of power.

In conclusion, hypothesis testing provides a rigorous framework for evaluating economic theories and drawing inferences from data. Understanding the concepts of null and alternative hypotheses, P-values, significance levels, Type I and Type II errors, and power is essential for conducting sound econometric research.

Regression Analysis: Unveiling Relationships Between Variables

Hypothesis Testing: The Foundation of Econometric Inference

Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic theories. By employing sophisticated statistical techniques, regression analysis emerges as a central method for uncovering and quantifying the relationships between economic variables, enabling researchers to test hypotheses and inform policy decisions.

The Core of Regression Analysis

Regression analysis serves as a cornerstone in econometrics, offering a structured approach to understand how changes in one or more independent variables are associated with changes in a dependent variable. It provides a framework for estimating the magnitude and direction of these relationships, while also assessing the statistical significance of the findings.

Types of Regression Models

Linear Regression

The linear regression model is perhaps the most widely used type of regression analysis. It assumes a linear relationship between the independent and dependent variables. It is particularly useful for examining the impact of one or more predictors on a continuous outcome variable. For example, a linear regression model might be used to assess the relationship between advertising expenditure and sales revenue.

Multiple Regression

Multiple regression extends the linear regression model to incorporate multiple independent variables. This allows for the simultaneous examination of the effects of several predictors on a single dependent variable. This is critical in economic contexts where outcomes are often influenced by a multitude of factors. For instance, predicting housing prices might involve considering factors such as square footage, location, number of bedrooms, and interest rates.

Non-Linear Regression

In many economic scenarios, the relationship between variables is not linear. Non-linear regression models are used to capture these more complex associations. These models can take various forms, including polynomial regression, exponential regression, and logarithmic regression, depending on the nature of the relationship being investigated. An example might be modeling the diminishing returns to scale in production, where the increase in output slows as input increases.

Ordinary Least Squares (OLS): A Fundamental Estimation Method

Principles of OLS

Ordinary Least Squares (OLS) is a fundamental estimation technique used in regression analysis. The primary objective of OLS is to minimize the sum of the squared differences between the observed values of the dependent variable and the values predicted by the regression model.

This minimization process yields estimates of the regression coefficients that provide the best fit to the data, under certain assumptions about the error term.

Assumptions and Limitations

The validity of OLS estimates relies on several key assumptions, including linearity, independence of errors, homoscedasticity (constant variance of errors), and no multicollinearity (high correlation among independent variables). Violations of these assumptions can lead to biased or inefficient estimates, necessitating the use of alternative estimation techniques.

Generalized Method of Moments (GMM): A Versatile Approach

Addressing Endogeneity and Misspecification

The Generalized Method of Moments (GMM) offers a more flexible estimation framework compared to OLS, particularly when dealing with endogeneity (correlation between independent variables and the error term) or model misspecification.

GMM allows researchers to specify a set of moment conditions that the parameters of the model must satisfy. It then estimates the parameters by minimizing a weighted distance between the sample moments and the theoretical moments implied by the model.

Advantages and Applications

GMM is particularly useful in situations where instrumental variables are available to address endogeneity. It can also accommodate various forms of model misspecification, making it a robust choice for complex econometric analyses.

For instance, GMM can be used to estimate the effects of government policies on economic outcomes, even when those policies are endogenous (i.e., correlated with unobserved factors that also affect the outcome).

Quasi-Experimental Methods: Approximating True Experiments

Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic models. While randomized controlled trials (RCTs) are the gold standard for establishing causal relationships, ethical, practical, or logistical constraints often preclude their implementation in economic research. This necessitates the use of quasi-experimental methods, which offer valuable alternatives for approximating true experiments and drawing causal inferences in real-world settings.

The Need for Quasi-Experimental Designs

In many economic scenarios, researchers lack the ability to randomly assign individuals or entities to treatment and control groups. This limitation arises from various factors, including ethical considerations, policy mandates, and the nature of the intervention being studied.

For instance, it is often impossible to randomly assign individuals to different educational programs or to manipulate government policies for research purposes. In such cases, quasi-experimental methods provide a framework for analyzing observational data and estimating treatment effects, albeit with careful consideration of potential biases and limitations.

Difference-in-Differences (DID)

Difference-in-Differences (DID) is a widely used quasi-experimental technique for evaluating the impact of a treatment or policy intervention. DID leverages the presence of both a treatment group, which is exposed to the intervention, and a control group, which is not.

The core idea behind DID is to compare the change in outcomes over time between the treatment and control groups. This approach effectively controls for pre-existing differences between the groups and accounts for common trends that may affect both groups equally.

Implementing DID

The DID estimator is typically implemented using a regression framework.

The dependent variable is the outcome of interest, and the independent variables include a treatment indicator, a time indicator, and an interaction term between the treatment and time indicators. The coefficient on the interaction term represents the estimated treatment effect.

Assumptions of DID

The validity of the DID estimator relies on the parallel trends assumption. This assumption states that, in the absence of the treatment, the treatment and control groups would have followed parallel trends in the outcome variable.

Violations of this assumption can lead to biased estimates of the treatment effect.

Advantages of DID

DID is relatively straightforward to implement and interpret.

It can control for both time-invariant confounders and common trends.

It provides a transparent and intuitive way to estimate treatment effects.

Limitations of DID

The parallel trends assumption can be difficult to verify empirically.

The presence of differential trends or time-varying confounders can bias the results.

DID may not be suitable for situations where the treatment effect varies substantially across individuals or groups.

Regression Discontinuity Design (RDD)

Regression Discontinuity Design (RDD) is another powerful quasi-experimental method that exploits sharp discontinuities in treatment assignment. RDD is applicable when treatment assignment is determined by whether an individual's value of a specific variable (the running variable) falls above or below a predetermined threshold.

For example, eligibility for a scholarship program may be determined by a student's score on an entrance exam, with a specific cutoff score determining eligibility.

Sharp vs. Fuzzy RDD

There are two main types of RDD: sharp RDD and fuzzy RDD. In sharp RDD, treatment assignment is a deterministic function of the running variable. Anyone above the threshold receives the treatment, and anyone below does not. In fuzzy RDD, treatment assignment is not deterministic. Some individuals above the threshold may not receive the treatment, and some below may receive it.

Implementing RDD

RDD typically involves estimating separate regression models for observations above and below the threshold. The difference in the predicted outcomes at the threshold provides an estimate of the treatment effect.

Assumptions of RDD

The key assumption underlying RDD is that individuals just above and below the threshold are similar in all respects other than their treatment status. This assumption, often referred to as local randomness, implies that the threshold acts as a quasi-random assignment mechanism.

Advantages of RDD

RDD can provide a credible estimate of the treatment effect in situations where the assignment mechanism is well-defined and understood.

It avoids the need for a control group by comparing individuals on either side of the threshold.

It has strong internal validity when the assumptions are met.

Limitations of RDD

RDD estimates a local average treatment effect (LATE), which may not be generalizable to other populations or settings.

It requires a precise understanding of the assignment mechanism and the running variable.

The treatment effect is only identified at the threshold, limiting the scope of inference.

Quasi-experimental methods, such as Difference-in-Differences and Regression Discontinuity Design, offer valuable tools for estimating treatment effects in situations where true experiments are not feasible. While these methods rely on specific assumptions and have their own limitations, they provide a rigorous framework for drawing causal inferences from observational data, contributing to informed decision-making and policy evaluation. Careful consideration of the underlying assumptions, potential biases, and the specific context of the research question is crucial for ensuring the validity and reliability of the findings.

Addressing Common Econometric Challenges: Ensuring Robustness

Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic models. While randomized controlled trials (RCTs) are the gold standard for causal inference, they are often impractical or unethical in economic contexts. Consequently, econometricians must grapple with a range of challenges to ensure the robustness and reliability of their findings. This section explores some of the most pervasive of these challenges and outlines strategies for mitigating their impact.

The Peril of Endogeneity

Endogeneity, a pervasive issue in econometric analysis, arises when an explanatory variable is correlated with the error term in a regression model. This correlation violates a core assumption of Ordinary Least Squares (OLS) estimation, leading to biased and inconsistent estimates.

There are several sources of endogeneity, including:

• Omitted variable bias
• Simultaneous causality
• Measurement error

Addressing endogeneity is crucial for obtaining valid causal inferences. Several techniques can be employed, including:

• Instrumental Variables (IV): This method involves finding a variable (the instrument) that is correlated with the endogenous regressor but uncorrelated with the error term.
• Two-Stage Least Squares (2SLS): A common implementation of IV estimation.
• Control Functions: Involves explicitly modeling the relationship between the endogenous variable and the error term.

The choice of technique depends on the specific context and the availability of suitable instruments or control variables.

Model Specification: Art and Science

Proper model specification is paramount to ensuring the validity of econometric results. A misspecified model can lead to biased estimates and misleading conclusions. This includes selecting the appropriate variables and functional form.

Variable Selection: The inclusion of irrelevant variables can reduce the precision of estimates, while omitting relevant variables can lead to omitted variable bias. Careful consideration of economic theory and prior empirical evidence is essential for guiding variable selection.

Functional Form: Choosing the correct functional form (e.g., linear, logarithmic, quadratic) is also critical. Visual inspection of the data, residual analysis, and the use of specification tests can help to determine the appropriate functional form.

Model Validation: Checking for Accuracy and Reliability

Once a model has been specified and estimated, it is crucial to validate its accuracy and reliability. Model validation involves assessing how well the model fits the data and whether its assumptions are met.

Several techniques can be employed, including:

• Residual Analysis: Examining the residuals for patterns that suggest violations of the model's assumptions.
• Goodness-of-Fit Tests: Assessing how well the model fits the data using measures such as R-squared or adjusted R-squared.
• Out-of-Sample Prediction: Evaluating the model's ability to predict outcomes on data that were not used in the estimation.
• Sensitivity Analysis: Assessing how sensitive the model's results are to changes in the specification or data.

The Importance of Clustered Standard Errors

In many econometric applications, data are clustered within groups (e.g., individuals within households, students within schools).

When error terms are correlated within these groups, the standard errors from OLS regression can be underestimated, leading to inflated t-statistics and an increased risk of Type I errors.

Clustered standard errors account for this within-group correlation, providing more accurate estimates of the standard errors and reducing the risk of false positives. It is crucial to use clustered standard errors when analyzing clustered data.

In conclusion, addressing these common econometric challenges is essential for ensuring the robustness and reliability of research findings. By carefully considering issues such as endogeneity, model specification, model validation, and clustered standard errors, econometricians can enhance the credibility and validity of their work.

Causal Inference: Establishing Cause-and-Effect Relationships

Addressing Common Econometric Challenges: Ensuring Robustness Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic models. While randomized controlled trials (RCTs) are the g...

While econometric analysis encompasses a broad range of methodologies, its ultimate objective frequently centers on uncovering and validating causal relationships between economic variables. Moving beyond mere correlation, establishing causality is paramount for informing policy decisions and understanding the true impact of interventions. However, the path to causal inference is fraught with challenges, demanding rigorous methods and careful interpretation.

The Centrality of Causal Relationships

The pursuit of causal inference lies at the heart of much of econometric research. Simply observing a statistical association between two variables is insufficient to conclude that one causes the other. Spurious correlations, confounding factors, and reverse causality can all lead to misleading inferences.

Therefore, econometrics provides a toolkit of techniques designed to isolate and identify genuine causal effects. This pursuit is crucial for evidence-based policymaking. If a policy intervention is found to have a causal impact on a desired outcome, then policymakers can be more confident in its effectiveness.

Instrumental Variables: Leveraging Exogenous Variation

Instrumental Variables (IV) estimation is a powerful technique for addressing endogeneity, a common obstacle to causal inference. Endogeneity arises when the explanatory variable of interest is correlated with the error term in the regression model, leading to biased estimates.

The IV approach involves finding an instrument, a variable that is correlated with the endogenous explanatory variable but uncorrelated with the error term. This instrument provides an exogenous source of variation in the explanatory variable, allowing for the identification of its causal effect on the outcome variable.

To be a valid instrument, the variable must satisfy two key conditions:

• Relevance: The instrument must be strongly correlated with the endogenous explanatory variable.

• Exclusion Restriction: The instrument must affect the outcome variable only through its effect on the endogenous explanatory variable.

Finding a suitable instrument can be challenging, and the validity of the IV estimates depends critically on the credibility of the exclusion restriction.

Control Functions: Addressing Omitted Variable Bias

Control function approaches offer another strategy for mitigating endogeneity and establishing causality. These methods involve explicitly modeling the relationship between the endogenous explanatory variable and the error term.

By including a control function (typically the estimated error term from the first-stage regression) in the main regression equation, we can account for the correlation between the explanatory variable and the error term. This allows us to obtain consistent estimates of the causal effect.

Control functions are particularly useful when a valid instrument is difficult to find. They are based on the assumption that the endogeneity arises from omitted variables. They rely on modeling the relationship between the endogenous variable and those unobserved factors.

Challenges and Limitations in Establishing Causality

Despite the availability of sophisticated techniques, establishing causality in econometrics remains a formidable task.

• Assumptions Matter: All causal inference methods rely on underlying assumptions, and the validity of the results depends on the plausibility of these assumptions. Researchers must carefully justify their assumptions and conduct sensitivity analyses to assess the robustness of their findings.

  You also like

  How to Annotate a Poem: Step-by-Step Guide

• Data Quality: The quality of the data is crucial for causal inference. Measurement error, missing data, and selection bias can all undermine the validity of the results.

• External Validity: Even if a causal effect is successfully identified in a particular setting, it may not generalize to other contexts. Researchers should consider the external validity of their findings and carefully assess the potential for heterogeneity in treatment effects.

• Complexity: Real-world economic phenomena are often complex and multifaceted. Isolating the causal effect of a single variable can be extremely difficult, requiring careful consideration of potential confounding factors and feedback loops.

In conclusion, establishing cause-and-effect relationships is a central goal of econometric analysis. While methods like instrumental variables and control functions provide powerful tools for addressing endogeneity, researchers must remain vigilant about the challenges and limitations inherent in causal inference. A critical approach, combined with careful attention to assumptions, data quality, and external validity, is essential for drawing meaningful conclusions.

Data Integrity and Research Ethics: Upholding Scientific Standards

Causal Inference: Establishing Cause-and-Effect Relationships Addressing Common Econometric Challenges: Ensuring Robustness Econometrics stands as the crucial intersection where economic theory meets statistical rigor. It is the application of statistical methods to economic data with the primary aim of giving empirical substance to abstract economic models. However, the value of econometric analysis hinges not only on the sophistication of the methods employed, but also on the integrity of the data used and the ethical conduct of the research process.

The Imperative of Data Integrity and Reproducibility

Data integrity refers to the accuracy, consistency, and reliability of the data throughout its lifecycle. In econometric research, this is paramount. Flawed or compromised data can lead to biased results, misleading conclusions, and ultimately, flawed policy recommendations.

Reproducibility, closely linked to data integrity, ensures that other researchers can independently verify the findings of a study using the same data and methods. Reproducibility is not merely a desirable attribute; it is a cornerstone of scientific credibility.

Without it, research loses its ability to inform policy and advance knowledge. Achieving reproducibility requires meticulous documentation of data sources, cleaning procedures, and analytical steps.

Transparency in Methodology and Data Usage

Transparency is essential for fostering trust in econometric research. Researchers must be forthright about the methodologies they employ, providing clear and comprehensive descriptions of their analytical techniques. This includes specifying the models used, the assumptions made, and any limitations encountered.

Similarly, the origins and characteristics of the data must be transparently disclosed. This involves identifying the data sources, explaining the sampling methods, and acknowledging any potential biases or limitations inherent in the data.

Furthermore, researchers should make their data and code publicly available whenever possible, allowing others to scrutinize and build upon their work. This practice not only enhances the credibility of the research, but also promotes collaboration and innovation within the field.

Ethical Considerations in Econometric Research

Ethical considerations permeate every stage of the econometric research process, from data collection to dissemination. Researchers have a responsibility to collect data in a manner that respects the privacy and autonomy of individuals or organizations involved. This may involve obtaining informed consent, protecting confidential information, and adhering to ethical guidelines established by relevant institutions or professional organizations.

During analysis, researchers must avoid manipulating data or selectively reporting results to support predetermined conclusions. It is essential to present findings objectively and transparently, even if they contradict expectations or challenge established theories.

Finally, in reporting research findings, researchers must acknowledge the limitations of their study and avoid overstating the implications of their results. Proper attribution of sources and avoidance of plagiarism are also crucial aspects of ethical research conduct.

Best Practices for Reliable and Credible Research

Upholding data integrity and research ethics requires a commitment to best practices throughout the research process. This includes:

• Implementing robust data validation procedures to detect and correct errors in the data.
• Maintaining detailed documentation of data sources, cleaning steps, and analytical methods.
• Conducting thorough sensitivity analyses to assess the robustness of findings to alternative assumptions or specifications.
• Seeking peer review and feedback from other researchers to identify potential weaknesses or biases in the analysis.
• Adhering to established ethical guidelines and institutional review board (IRB) requirements.
• Making data and code publicly available whenever possible to promote reproducibility and transparency.

By adhering to these best practices, econometric researchers can ensure that their work is reliable, credible, and ethically sound. This, in turn, strengthens the foundation for informed decision-making and evidence-based policy.

So, there you have it! Hypothesis testing is a crucial part of the economist's toolkit. Whether it's running regressions, analyzing survey data, or constructing complex economic models, economists use a variety of methods to test a hypothesis and ultimately, to better understand the forces shaping our economy. Hopefully, this has given you a clearer picture of how they do it.