1: Introduction

Question 1

What are five concepts connected to Econometrics?

Answer 1

• Linear regressions
• Moderating effects
• Mediating effects
• Time series
• Panels

Question 2

What concepts are connected to machine learning?

Answer 2

• Supervised learning
• Unsupervised learning
  (- Reinforced learning)

Question 3

What are five concepts connected to research design?

Answer 3

• Data analytics process
• Research question
• Hypotheses
• Data collection
• BSc thesis

Question 4

What is solving problems in DA like?

Answer 4

Like solving them through a lens: how we frame the situation affects the results.

Question 5

What is a good quantitative research question?

Answer 5

• Clear
• Focused
• Allows the writer to take an arguable position but does not leave room for ambiguity.
• Avoid the “all-about” questions.
• Empirically addressable.
• Something that you are interested in or care about.

Question 6

What are six questions to ask for evaluating a research question?

Answer 6

• Is the research question clear?
• Is it too broad or too narrow?
• Is it something I/others care about?
• Is it a new spin on an old idea, or does it solve a problem?
• Is it researchable within the given time frame and location?
• Is it measurable and will the process produce data that can be supported or contradicted?

Question 7

Why is a research questions (RQ) important?

Answer 7

• Clarifies what you are looking for (e.g. core concepts).
• Determines your research method.
• Helps you avoid asking something that cannot be answered empirically.
• Narrows down your theory and data collection.
• Enables you to communicate to others involved (co-researchers, participants, audiences, etc).

Question 8

What is a quantitative research question?

Answer 8

Used to answer questions on relationships with an intention to explain, predict and control a phenomena.

Question 9

What is a qualitative research question?

Answer 9

Used to answer relatively open-ended questions on experiences, interpretations and meanings.

Question 10

What two overarching research methods are there?

Answer 10

• Quantitative methods
• Qualitative methods

Question 11

What are quantitative research methods?

Answer 11

• Numbers and anything that is measurable in a systematic way of investigation of phenomena and their relationships.
• Used to answer questions on relationships, with an intention to explain, predict, and control a phenomena.
• Functions: identify a cause, quantify the effect of a cause, predict outcomes, explain, test, and refine theory or hypothesis.

Question 12

What are qualitative research methods?

Answer 12

• Data like words and observations, which may be unstructured, and which often focus on meanings.
• Used to answer relatively open-ended questions on experiences, interpretations, or complex processes.
• Functions: Understand, process, explore, interpret, create foundation for theory.

Question 13

What is a theory?

Answer 13

A statement about HOW a set of constructs are related to each other explaining a broad phenomenon.

• Answers questions of “how”, “when” and “why”, as opposed to only “what”.
• An important purpose of a theory section is to ground hypotheses.

Question 14

What is a hypothesis?

Answer 14

A proposed explanation of a fairly narrow phenomena or set of observations.
- Not a guess: a theory-driven and informed attempt to describe what has been initially observed.

(statement that can be true or false)

Question 15

What is the role of hypotheses?

Answer 15

1. Identify and organize different issues that we need to consider in order to ANSWER THE RESEARCH QUESTION.
2. Provide statements of how different constructs in the model are linked together.

Question 16

A hypothesis is true or false, we decide to…

Answer 16

… accept (with uncertainty) or reject it

Question 17

What is a Type I error?

Answer 17

The null hypothesis is true, but we reject it.

Question 18

What is a Type II error?

Answer 18

The null hypothesis is false, but we accept it

Question 19

What is the connection between constructs and variables?

Answer 19

A construct/concept is an abstract category in your theory which cannot be measured directly. In empirical research, you assess constructs using variables, which can be measured.

Question 20

What is research design about?

Answer 20

HOW data are collected.
- Experimental
- Descriptive
- Correlational
Research question guides research design.

Research design is all the necessary steps around the data analytics for it to make any sense.

Question 21

What are some strengths of quantitative methods?

Answer 21

• Scientific objectivity: data can be interpreted with statistical analysis, and since statistics are based on the principles of mathematics, the approach is viewed as scientifically objective and rational.
• Most useful for testing, validating and refining already constructed theories.
• Fast analysis
• Replication

Question 22

What are some weaknesses of quantitative methods?

Answer 22

• Variability of data quantity
• Fairly narrow settings
• Be aware of three sources of biases: you, the respondent, and the method.

Question 23

Correlation is not causation. Which three conditions must be met for causality?

Answer 23

1. The cause must precede the effect.
2. The cause and effect should correlate.
3. All other explanations of the cause-effect relationship must be ruled out. An effect should be present when the cause is present, and when the cause is absent the effect should also be absent.

Question 24

What are the four main types types of quantitative research designs?

Answer 24

• Descriptive
• Correlational
• Quasi-experimental
• Experimental

Question 25

What is descriptive research design?

Answer 25

• Explore the characteristics of samples and population.
• Purpose: observe and document situation as it occurs (national bureau of statistics).
• Makes no hypothesis on relationship between variables, no manipulation, no random draws per groups.
• Use to define and characterise.

TYPICAL QUESTION: What is…?

Question 26

What is correlational research design?

Answer 26

• Explores relationship between variables using statistical analyses.
• Tries to study relationship between variables in a quantifiable way.
• Snapshot of many variables at single point in time, or repeated measures at different points in time.

TYPICAL QUESTION: What influences…?

Issue: Tertium quid, correlation research cannot demonstrate causality.

Question 27

What is quasi experimental research design?

Answer 27

• Focusing on cause-effect relationship among variables, without possibility to perform manipulation.
• Researcher does not randomly assign groups and might not manipulate the IV.
• Potential endogeneity is addressed through statistical manipulation: fixed effect, matching methods, Heckman corrections, etc.

TYPICAL QUESTION: What influences…?

Issue: Because research cannot manipulate causes one by one, one quasi experiment is not enough to prove any cause-effect relationship.

Question 28

What is experimental research design?

Answer 28

• Also called true experiment or lab experiment. Researcher manipulate the IV over the participants, with random assignment, and is therefore able to observe variation occurring after the only manipulation.
• Control for all variables except the one being manipulated (the IV).

TYPICAL QUESTION: What influences…?

Issue: Do not account for the context where the effect is supposed to occur. Effects may be observed in laboratory setting but become invisible in real/observed situations.

Question 29

What are four sources of observations?

Answer 29

• Measured
• Declared
• Direct observations
• Indirect observations

Question 30

What are control variables?

Answer 30

Other factors than our IV that can have an impact on our DV, so we need to control for them in the regression. Can be identified through previous work on the subject

Question 31

What is the Adjusted R2- value?

Answer 31

Shows the explanatory power of our model. It is adjusted for multicollinearity.