Lecture 1:

The research question consists of:

• Starts with the research question
• Theory/literature
• Hypothesis: testable prediction
• Then we can define the main variables (dependent and independent variable)
• Then you need to collect data (measurement)
• Analyze data (graphically/descriptively)
• Fit a model
• Conclusion

Measurment: a relationship between the numbers and what is being measured. You can measure variables in different kind of ways. Important to consider:

1. What do you really want to measure
2. What is your research question

Basic issues in measurement:

1. Validity: extent to which a measure correctly represents the concept of a study (refers to the study not a specific variable

• Internal validity:  how well the study was done
• External validity: generalize results to other situations

2. Accuracy: is the measure close to the actual value and did you get the right answer on average?
3. Reliability: extent to which a variable is consistent in what it is intended to measure

It is important to measure the right thing and be clear about what you measure

Organize your data:

• Cross sectional: observations at a given points or time
• Time series
• Panel: both cross sectional and time-series dimensions (over a period of time)

Article Hult et al:

• Focus on why do some firms outperform others
• Performance is important variable (often DV)
• Inconclusive results about determinants of performance
• Conclusions depends on the measurement of performance
• No systematic investigation as to how IB research measures performance (contribution)
• They examine the measurement of performance
• They do that in 96 articles published in the journal between 1995 and 2005
• More specifically: they asses the measurement of performance in 3 dimenstions:

1. Type of data source
2. Type of measure
3. Level of analysis

• What did they find: Most studies do not measure performance in a manner that captures the multifacted nature of the construct
• We describe the implications of these results and offer suggestions for improving future practice (present non binding guidelines)-what they do with their findings

Questions the researcher has to deal with:

• What do you want to measure
• What kind of data to use:

1. Primary data: collected by researcher à time consuming, but original
2. Secondary dataL collected by other agencies; cheap, but lacks originality and may not be fitting to the research question

• How should we measure performance, because you can measure the same thing in different ways

1. Financial performance: reflects economic goals
2. Operational performance: non financial, like innovation, productivity and satisfaction
3. Overall effectiveness: e.g. reputation (related to both)

• Which level of analysis to focus on (remember external validity):

1. Firm
2. Strategic business unit (SBU)
3. Inter-organizational unit (cannot be found in Hult et al)
4. In general at very different levels, country, region, industry, firm etc.

Why do all they matter: potentially different results and conclusions. The importance of the measurement

But: Is one right and the other wrong. Should we all agree on a single measure to use? So which one to choose? It depends on the research question

Tip:

• Be clear about what you want to do
• Choose the appropriate measure for your analysis
• Justify your decision
• Be clear about you limitations

Selection bias:

• Be careful about the interpretation of your result and be careful about what conclusion you draw.
• Can compare with another sample

Endogeneity:

• Correlation between regressor and error term
• Reasons: measurement error, omitted variable (any variable that is not included as dependent variable but could influence the dependent variable) and reverse causality (causality that is not really causality)
• Standard OLS estimate biased
• There are solutions

Lecture 3:
research process:

• Research questions, motivations and contribution
• Hypotheses
• Data
• Analysis
• Results
• Check whether the results confirm hypothesis

Correlation:

• Correlation coefficient r
• Bounded measure

Regression analysis:

1. simple bivariate regression: one single variable

• x1: explanatory variable; independent variable; regressor; predictor variable …
• y: dependent variable and outcome variable
• B0: intercept/constant
• B1: slope; regression coefficient for the variable x1 (size or magnitude of the effect x1 on y)
• E: error term
• We want to estimate B0, B1 ( you have estimated veriables

How to estimate:

• OLS: technique to best represent the linear relationship between x  and y (minimize the sum of standard deviations form the line) (minimize the sum of the squared residuals)
• How to draw a straight line: Find the one that best fits the data (goes through, or as close to, as many of the data points as possible)
• Minimize sum of the estimated errors

2. Multiple regression: two or more independent variables: we analyze the relation relationship between a single dependent variable and several independent variables

• The coefficient of each independent variable indicates change in dependent variable when the independent variable changes but all other independent variables remain constant (ceteris parisbus)

Dummy variables independent variables:

How many dummies to include:

With intercept include m-1 dummies the omitted one is the base/reference,  However if you have only two categories one dummy is sufficient to represent both. Otherwise, one is perfectly predicted  by the other (exact linear relationship) and the regression coefficients cannot be estimated.

Regression coefficients:

What to look for when interpreting:

1. Sign
2. Significance
3. Size, magnitude
4. Then interpret

• The regression coefficients b1 coefficient measures the amount of change in y due to change x1 while the other regressors are held constant (ceteris paribus)
• It represents the change in the DV for a one-unit increase in the IV while the other regressors are held constant.

Article:

What is about:

• LMEs vs CMEs
• Flexible vs rigid labour markets.
• What does this mean for management bureaucracies: starting point

Hypothesis:

• High shares of flexible workers à high shares of managers
• Labour market flexibility à lack of trust, loyalty, commitment à management and control required
• Labour market flexibility à Lack of learning and knowledge accumulation à bad for innovation and labour productivity growth

Why are we doing this:

• Insignificant or conflicting results in literature
• Sparse empirical work on topic
• Important to test the theory at the firm level
• The Netherlands represent an interesting case to do so: CME but with LME features

What are we looking at:

• Data source: SCP organization-level survey data. For the Netherlands à Data collection: telephone interviews plus a postal survey.
• Information on:

1. Organizations
2. From various sectors
3. Percentage of employees with managerial positions in an organization
4. Shares of flexible workers
5. Industry
6. Firm size

• Survey of 2009-10 is used

Theory à variables à data à model

• The substantial variation in shares of flexible workers (IV à combination of % of temporary workers and %  of manpower agency workers plus freelancers) allows analysing the impact of the latter on management rations (DV), controlling for a number of influential factors ( Fim size, growth, age, dummies for: restructuring operation; competitive market; sensitive to business fluctuations; industry dummies. How am I Supposed to know? à look at literature

A note on the regression equation:

• The hypothesis lead to the regression equation
• Sometimes the regression equation is explicitly presented in the paper
• Here, the regression equation is not presented, but you should be able to infer it from the text and/or regression table
• Choose key IV and then look at regression coefficient. Also look at significant level

Assessing your model:

• Goodness fit
• Coefficient of determination (R-squared):

1. Proportion of the variation in y that is explained by the linear combination of the x variables
2. Between 0 (no prediction) and 1 (perfect prediction)
3. In a bivariate linear regression, it equals the squared correlation coefficient
4. If r2 between 0.21 and .30, 30% of the variation in y is explained by the x’s

• R2 increases as we keep adding variables
• Adjusted coefficient of determination: Adjusted r2:

1. Modified measure of R2 that takes into account the number of independent variables and sample size
2. Some punishment for the inclusion of additional explanatory variables
3. When comparing between models, search for highest r2

• F test: A test for the overall significance of the model, tests whether all parameters are jointly zero b1=b2à compare to critical values of F-distribution or look at corresponding p-value or significance

Estimates and significance:

• Does my main IV have a significant impact on the DV
• How can I determine this: Statistical significantce (Compare to 0 and t-test)
• Note: statistical vs economic significane
• First formulate Null and alternative hypothesis H0 bj=0 h1: bj is not 0
• Choose the significance level
• Option 1:

• Option 2:

•  So if p-value<0.01, the estimate is significantly different from 0

Lecture 5:

1. Ouliers: Data point that does not follow the general trend of the data (extreme value)

What does happen if we run the regression anyway?

• The fit of the model can change
• The regression may be titled
• You can remove an outlier

How can we check whether we have outliers

• Scatterplots
• Statistical test
• Easiest wat the range of +/- 2 à 3 standard deviatons include the word at least in the conclusion! à based on the assumption that it distribution is normal

How can we solve the problem?

• First think: what is the reason for the outlier, can/could you do something?
• Throw the outlier out of the dataset, however mismeasurement, error in the observation, data entry error. But not because it’s convenient to do so.
• Be careful: some extreme values are to be expected, indicative of the characteristics of the population. Therefore it is important to check how sensitive your results are to the presence of the outlier? à what happens if we keep the outlier, what happens if we omit the outlier.
• If the outlier does not change the results, but does affect assumptions, you may drop the outlier
• If it affects both results and assumptions, you may not drop the outlier, but you have to run the regression both with and without the outlier and say that in he paper
• If a relationship is clearly created by the outlier, you may drop the outlier, because without it there would be no relationship between x and y. So the regression coefficient does not truly describe the effect of x on y

2. Reverse causality: We assume that changes in the dependent variables are caused by changes in the independent variables. But we only find a statistical relationship, says nothing about causality of the direction of causality. In some analysis is could be that y (also) causes X which is called reverse causality à cause endogeneity problem

How can we check whether there is a reverse causality problem?

• What does the theory say
• Timing of measurement: Theory says x causes y, but sometimes x is measured later than y
• Statistical tests (to check whether changes in x precede changes in y) and some more advances techniques

What to do:

• Have a model that is well-grounded in theory
• Explain
• Acknowledge
• In general: advances econometric techniques also exist to mitigate the problem of endogeneity

3. Omitted variable bias: which variables to include as IVs and what happens if we omit relevant variables? You have omitted variables if:

• As excluded variable has some effect on your DV and
• It’s correlated with at least one of your IVs (endogeneity)

It is impossible to control for everything, so how do we solve the problem?

• Avoid simple regressions models (with on IV)
• Include variables that are likely to be the most important theoretically in explaining the DV (what does the literature say)

Panel data or longitudinal data: data on many units collected at several points in time, whereby each unit is observed several times. You also have cross sectional and time series dimensions.

Why panel data:

1. Rich in information
2. Potentially, an increase in sample size
3. Possibility to control for time-invariant effects correlated with the regressors
4. How> intuition: Include dummy variables for each cross-section unit and use fixed effects.
5. Mitigate omitted variable bias

Fixed effects model: is a statistical regression model in which the intercept of the regression model in which the intercept of the regression model is allowed to vary freely across individuals or groups. It often applied to panel data in order to control for any individual-specific attributes that do not vary across time. Remove omitted variable bias. Assumption: the individual-specific effects are correlated with the IV’s

Assume: For the Grundfeld data we concluded that the assumption of OLS regression that the investment behaviour of all firms in all years is the same à is not realistic. The fixed effects model offers another way of restricting that assumption, namely by assuming that each firm has a number of unique characteristics that influence the firm’s investment behaviour. These unique characteristics are caught in the model by including for each firm a separate dummy variable.

In example of Grunfeld we assume:

• Each firm has a unique characteristic which is stable over time
• Random error term is assumed to satisfy the usual OLS assumptions
• Hence each firm I gets a different intercept parameter but the slope coefficient b2 and b3 are assumed to be the same for all firms
• An easy way to estimate the model is to create for each firm a dummy variable and add thse dummies to the model

General equation FE model

Restrictions of FE model, the FE model is very powerful but:

• We cannot include variables that do not vary over time, all stable characteristics are captured by dummies, it leaves not variation left for estimating effects of variables that vary between economic entities
• You can only include those that change over time
• However you can still examine the interaction between group dummies and time-varying variables in FE model

When should you use a Fe model à if you are concerned about omitted factors that may be correlated with key predictors at the group level

Interpretation of results à similar to OLS

Logs in the regression equation, in general don’t forget:

• Sign-size significance
• Use the unit of measurement of y and x when given
• Ceteris paribus

4 situations:

Robustness/sensitivity analysis:

To what end? à determine how sensitive your results are to change in the model

Experiment with:

• Combinations of (other) control variables
• Datasets
• Time frames

Always rely on theory and literature

Do you results remain, results are robust.