An advanced-level thesis in Applied Statistics at Örebro University School of Business, dated Autumn 2014. It develops and demonstrates a simulation procedure for estimating the size and power of the Hausman test for choosing between fixed-effects and random-effects panel models. Using Grunfeld investment data for 10 companies over 1935 - 1954, the study compares Swamy - Arora, Amemiya, and Nerlove random-effects estimators with asymptotic, bootstrap, and Monte Carlo critical values where feasible. In the studied settings, asymptotic critical values produce over-rejection with Swamy - Arora and Amemiya estimation and severe under-rejection with Nerlove estimation. Bootstrap critical values improve size accuracy, while Monte Carlo critical values yield sizes close to nominal levels within the simulated designs. Test power remains low, particularly at small significance levels. The thesis includes methodology, results, recommendations, and a detailed simulation algorithm.
Free preview text
The first 5 pages, shown as text so search engines and readers can find this document.
Page 1
Örebro University
Örebro University School of Business
Master's program "Applied Statistics"
Advanced level thesis I
Supervisor: Prof. Panagiotis Mantalos
Examiner: Sune Karlsson
Autumn 2014
The Accuracy of the Hausman Test in Panel Data:
a Monte Carlo Study
Teodora Sheytanova
90/04/04
Page 2
1
Abstract
The accuracy of the Hausman test is an important issue in panel data analysis. A procedure
for estimating the properties of the test, when dealing with specific data, is suggested and
implemented. Based on simulation that mimics the original data, the size and power of
Hausman test is obtained.
The procedure is applied for different methods of estimating the panel data model with
random effects: Swamy and Arora (1972), Amemiya (1971) and Nerlove (1971). Also, three
types of critical values of the Hausman statistics distribution are used, where possible:
asymptotical and Bootstrap (based on simulation and bootstrapping) critical values as well
as Monte Carlo (based on pure simulation) critical values for estimating the small sample
properties of Hausman test.
The simulation mimics the original data as close as possible in order to make inferences
specifically for the data at hand, but controls the correlation between one of the variables
and the individual-specific component in the panel data model.
The results indicate that Hausman test over-rejects the null hypothesis if performed based
on its asymptotical critical values, when Swamy and Arora and Amemiya methods are used
for estimating the random effects model. The Nerlove method of estimation leads to
extreme under-rejection of the null-hypothesis. The Bootstrap critical values are more
appropriate. With the example data used, the chosen bootstrap procedure and the specific
number of bootstrap samples, the bootstrap size tends to follow the upper limit of the
confidence interval of the nominal size, although sometimes it passes the limit line and a
slight over-rejection is observed. The simulations show that the use of Monte Carlo critical
values leads to an actual size very close to the nominal.
The power of Hausman test proved to be considerably low at least when a constant term is
used in the modelling.
Keywords: Hausman Test, Panel Data, Random Effects, Fixed Effects, Monte Carlo,
Bootstrap.
Page 3
2
Contents
1. Introduction .................................................................................................................................... 4
1.1. Background ............................................................................................................................. 4
1.2. Statement of the problem and purpose of the research ........................................................ 4
1.3. Organisation of the thesis ....................................................................................................... 5
2. Theoretical basis ............................................................................................................................. 5
2.1. Panel data ............................................................................................................................... 5
2.2. Estimation methods ................................................................................................................ 6
2.2.1. Pooled model .................................................................................................................. 7
2.2.2. Fixed effects model ......................................................................................................... 8
2.2.3. Random effects model .................................................................................................... 8
2.3. Hausman test for endogeneity – general case ..................................................................... 10
2.4. Hausman test for differentiating between fixed effects model and random effects model 11
2.5. Simulation and resampling methods .................................................................................... 12
2.5.1. Monte Carlo simulations ............................................................................................... 12
2.5.2. Bootstrap....................................................................................................................... 12
3. Methodology: Design of the Monte Carlo simulation .................................................................. 13
3.1. Defining a set of data generated processes (DGP). .............................................................. 13
3.2. Generating r independent replications................................................................................. 16
3.3. Generating bootstrap samples in order to obtain Hausman test critical values .................. 16
3.4. Obtaining Hausman test critical values through the use of a pure Monte Carlo simulation 19
3.5. Performing Hausman test in each replication ...................................................................... 19
3.6. Estimating the size and power .............................................................................................. 19
3.7. Note ...................................................................................................................................... 20
4. Implementation: Reporting the Monte Carlo experiment ........................................................... 21
4.1. Data ....................................................................................................................................... 21
4.2. Technical details.................................................................................................................... 21
4.3. The model of the data generated process ............................................................................ 22
4.4. Obtaining Hausman statistics critical values trough bootstrapping ..................................... 23
4.5. Obtaining Hausman statistics critical values through pure Monte Carlo simulation ........... 24
4.6. Results ................................................................................................................................... 24
4.6.1. Size and power of Hausman test (Amemiya method for estimating random effects
model) 25
4.6.2. Size and power of Hausman test (Nerlove method for estimating random effects
model) 33
Page 4
3
4.6.3. Size and power of Hausman test (Swamy and Arora method for estimating random
effects model) ............................................................................................................................... 40
4.7. Discussion.............................................................................................................................. 43
4.8. Recommendations ................................................................................................................ 44
5. Summary and Conclusion ............................................................................................................. 45
5.1. Summary ............................................................................................................................... 45
5.2. Conclusion ............................................................................................................................. 45
6. Bibliography .................................................................................................................................. 46
Appendix ............................................................................................................................................... 48
6.1. Fixed effects estimation methods......................................................................................... 48
6.1.1. Within-group method ................................................................................................... 48
6.1.2. Least squares dummy variable (LSDV) method ............................................................ 49
6.2. Algorithm .............................................................................................................................. 49
Page 5
4
1. Introduction
1.1. Background
When performing a statistical hypothesis test an issue that must be considered is the
accuracy of the test. There are two properties that define the accuracy of a hypothesis test:
its size and power. The size is the probability of rejecting the null hypothesis, when it is the
correct one and in social sciences tests are usually run at significance level 5%, which
guarantees that if the null hypothesis is correct and a number of tests are made based on
different samples of the same population, in 95% of the cases the null hypothesis won’t be
rejected. The power represents the probability of correctly rejecting the null hypothesis.
Values of the power of 80% or above are considered “good” when corresponding to size of
5% (Cohen, 1988).
An issue arises of whether setting the significance level to a particular value will actually
result in a risk of making a Ist type error of the same value. Based on a particular sample it
might turn out that the test has bigger (or smaller) size than what is set initially. It is
reasonable to conduct a research for finding the true value of the size of the performed test
as well as of its power.
The focus on this thesis is Hausman test, used for choosing between models in panel data
studies. Hausman test examines the presence of endogeneity in the panel model. The use of
panel data gives considerable advantages over only cross-sectional or time series data, but
the specification of the model to be used is of great importance for obtaining consistent
results. One of the tests used to determine an appropriate model is Hausman test, which
specifies whether fixed or random effects panel model should be used. As one of the most
used tests in panel data analysis, the study of its properties should represent a great
interest.
However, not many publications have been made in this sense. Many research papers like
Wong (1996), Herwartz and Neumann (2007), Bole and Rebec (2013) focus on bootstrapping
Hausman test in order to improve its finite sample properties and find the true distribution
of the test, but there are less research efforts done on the subject of Hausman test’s
properties in specific. Jeong & Yoon (2007) explore the effect of instrumental variables on
the performance of Hausman test by simulating data.
This thesis illustrates the estimation of Hausman test’s size and power procedure, which can
be implemented for a particular real data study. By applying the methods used in this thesis,
one can examine the accuracy of Hausman test in a particular case for a specific research
and not rely on general studies based on data, which might not fit their individual instance.
1.2. Statement of the problem and purpose of the research
The thesis illustrates a procedure that can be used for obtaining the actual size and power of
the Hausman test in a particular study, for a specific sample. The research has been
performed on real data and the methodology, which has been used, can be applied for
different panel data cases.
The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study: summary, contents and background
This guide helps you understand the scope of The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study before or alongside reading the original document. It is based on the catalogue record, credited creator, source, classification and topics shown on this page. It does not replace the original work and does not invent claims about passages that Doxxly cannot verify.
Summary and overview
The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study is a 54-page documentcredited to Umar P. Ingleby. An advanced-level thesis in Applied Statistics at Örebro University School of Business, dated Autumn 2014. It develops and demonstrates a simulation procedure for estimating the size and power of the Hausman test for choosing between fixed-effects and random-effects panel models. Using Grunfeld investment data for 10 companies over 1935 - 1954, the study compares Swamy - Arora, Amemiya, and Nerlove random-effects estimators with asymptotic, bootstrap, and Monte Carlo critical values where feasible. In the studied settings, asymptotic critical values produce over-rejection with Swamy - Arora and Amemiya estimation and severe under-rejection with Nerlove estimation. Bootstrap critical values improve size accuracy, while Monte Carlo critical values yield sizes close to nominal levels within the simulated designs. Test power remains low, particularly at small significance levels. The thesis includes methodology, results, recommendations, and a detailed simulation algorithm. It appears in the Business & Economics collection, which groups it with other material that readers may use for comparison and further study.
The item can be approached in several ways. A reader looking for a quick orientation can begin with the free five-page preview, inspect the page structure and use the metadata to judge relevance. Someone undertaking deeper research can work through the complete file, search its indexed text, record useful page references and compare its treatment of Hausman test, Panel data, Random effects, Fixed effects, Monte Carlo simulation, Bootstrap, Statistical power, Applied statistics, Master's thesis with the related resources linked below.
Rather than reducing the work to a single claim, this page treats it as a source that should be read in context. The title, creator, publication source, licence and format all help explain what kind of evidence it may offer. Readers should distinguish the original author's statements from Doxxly's catalogue description, and should cite the original edition when using the material in academic, professional or published work.
Key topics and reader takeaways
The catalogue connects this document with Hausman test, Panel data, Random effects, Fixed effects, Monte Carlo simulation, Bootstrap, Statistical power, Applied statistics, Master's thesis. These terms offer practical entry points rather than a substitute for close reading. Start by identifying the definitions, examples, arguments, tables or visual explanations attached to each topic. Then note where the creator draws connections, makes distinctions or moves from background material to a conclusion, recommendation, calculation or narrative development.
A useful reading method is to separate description from interpretation. First record what the item actually presents: names, headings, figures, formulas, quotations or stated propositions. Next consider why those details matter within the whole work. Finally compare them with another source from the same subject area. This process helps turn a long digital file into evidence that can be evaluated rather than merely skimmed.
Table of contents and reading guide
The following is an editorial content guide based on the verified catalogue topics. It is not represented as the original author's formal table of contents. Use it to organise notes while moving through the 54 pages in the player.
Part 1
Hausman test
Use this section of the reading guide to identify how Hausman test appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 2
Panel data
Use this section of the reading guide to identify how Panel data appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 3
Random effects
Use this section of the reading guide to identify how Random effects appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 4
Fixed effects
Use this section of the reading guide to identify how Fixed effects appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 5
Monte Carlo simulation
Use this section of the reading guide to identify how Monte Carlo simulation appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 6
Bootstrap
Use this section of the reading guide to identify how Bootstrap appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 7
Statistical power
Use this section of the reading guide to identify how Statistical power appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
Part 8
Applied statistics
Use this section of the reading guide to identify how Applied statistics appears in The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, note important terms, and connect the subject to the wider purpose of the work.
When a heading in the original differs from this guide, follow the original heading in your citation. Page numbers can also vary between editions. Doxxly's direct page field is useful for navigation, but formal references should include the title, creator, source and edition details available in the file itself.
About Umar P. Ingleby
Doxxly credits Umar P. Ingleby as the author or creator of The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study. This page does not add an unverified biography, job title or institutional affiliation. The attribution is preserved so readers can distinguish the creator from the hosting library and follow the original record when an external author link is available.
Author context matters because a work reflects a particular purpose, audience and point of view. Before relying on the material, consider what the creator set out to explain, when the source was produced and whether later research or revised guidance should also be consulted. For literature and philosophy, compare translations and editions; for technical material, check whether a newer version has superseded the one shown here.
About the source organisation
The recorded source for this item is diva-portal.org. Doxxly provides access and discovery tools but does not claim to be the original publisher unless explicitly stated. Source links, creator attribution and the recorded Publicly available file collected for editorial review licence are displayed to help readers assess provenance and permitted use.
Provenance is especially important for reports, standards, presentations and working spreadsheets. Check the original source for amendments, supporting files and current policy. A trustworthy research workflow keeps the source URL with any notes or downloaded copy, verifies important figures against the original, and avoids treating an archived document as automatically current.
Begin with the title and opening pages, then use the guide above to decide whether to read sequentially or move to a particular topic. The Doxxly player provides previous and next navigation, a direct page field, zoom steps, fit-to-width and fit-to-page modes, rotation and fullscreen. These controls are useful when a page contains small type, a wide table, an image, a scanned layout or a dense sequence of notes.
The first five pages are open to every visitor. A plan unlocks the complete item and full-document search. Search is best used for distinctive names, technical terms or short phrases; broad words can produce too many possible contexts. When you find a useful passage, read the surrounding pages so that a sentence, chart or formula is not separated from its explanation.
If downloading is enabled, members can obtain the original PDF file. Keep the original format when formulas, slide sequencing, pagination or layout carries meaning. Respect the Publicly available file collected for editorial review terms displayed on this page, and use the rights-report link if you believe the attribution or availability needs review.
Research questions to consider
As you read The Accuracy of the Hausman Test in Panel Data: a Monte Carlo Study, ask which audience the creator addresses, what problem or idea organises the work, what evidence is presented, and which assumptions remain unstated. Note whether important terms are defined consistently and whether examples represent a broader pattern or a narrow case. For quantitative material, record units, time periods, formulas and data sources. For interpretive material, separate the creator's language from your own summary.
A second set of questions concerns relevance today. Which parts remain durable, which depend on their historical or institutional setting, and what would need confirmation from a newer source? Following the tags and the Business & Economics collection can reveal useful comparisons. Agreement between sources can strengthen confidence, while disagreement can point to changing definitions, methods, evidence or priorities.
Finally, decide what you need from the item: background knowledge, a quotation, a worked example, a model, a visual explanation or a starting bibliography. That purpose should guide the depth of reading and the notes you keep. Doxxly helps locate and navigate the source; careful interpretation remains the reader's responsibility.