QRM III
Monday lecture
Wednesday lecture
Thursday ISC and Tutorial
Inductive reasoning: drawing a conclusion based on an observation
• Outcome of inductive reasoning is one or more hypotheses
Deductive reasoning: conclusion follows from a set of reasons (premises)
• Begins with a hypothesis
Possibility to combine these in research process.
More valid for deductive reasoning: research process
Theory → constructs/concepts → variables
Hypothesis: expected relationship between variables based on a theory.
- Descriptive hypothesis
o Value of one variable
o 80% of students pass the QRM course
- Rational hypothesis
o Relationship between 2 variables of one case
o Students who spend more time eon the QRM course get a higher grade
Theory: a set of interrelated concepts, definitions, and propositions to explain and predict
phenomena.
→ theory is not the opposite of a fact
Construct: conceived (by the researcher) specifically for the study or for the theory building
- Difficult to observe
- Often built from concepts
- E.g., yuppies, composed of concepts (age, type of car)
Dependent variable: the variable that we want to explain
Independent variable: variable that explains (in part) the dependent variable
Moderating variable: the strength of the relationship between a dependent and independent
variable is affected by a moderating variable
Intervening variable: the influence of an independent variable on a dependent variable is
through an intervening variable.
Hallmarks of scientific research
, • Purposiveness: the supervisor or the manager of an organization has started the
research for a specific purpose.
• Rigor: carefulness, honesty, and the degree of accuracy in research investigations.
• Testability: researcher develops a hypothesis on how for instance, employee
commitment can be improved
• Replicability: we repeat the test of hypotheses again and again for the better outcomes.
• Precision and confidence: precision reflects the degree of accuracy of the results based
on the sample to what really exists in the universe. Confidence refers to the probability
that out estimations are correct.
• Objectivity: data analysis should be based on the facts of the findings based on actual
data
• Generalizability: scope of applicability of the research findings in one organizational
setting to other settings.
• Parsimony: it can be introduced with a good understanding of the problem and the
important factors that influence it.
Descriptive research
- Inventory of level of absenteeism
- How often does something occur
- Reasons/possible causes why something occurs do not attend
Causal research
- Testing cause and effect relationships
- What factors lead to difference in absenteeism?
Descriptive research → inductive reasoning → exploratory
,Causal research → deductive reasoning → hypothesis testing
Exploratory vs formal (hypothesis testing)
Exploratory
- Looser design
- Aim is to develop theories that can be tested in further research
- Ends with the formulation of hypotheses
Formal (hypothesis testing)
- Begins with the hypotheses
- Established procedure for hypothesis testing
Role of the researcher
Contrived setting: artificially created environment
Non-contrived: a natural environment
• Manipulation
• Simulation
• Controlled environment
Intensive vs extensive design
Intensive design: Intensively investigate a case
• Case study. Invite 5 students for an interview
Extensive design: Investigate many cases less deeply
• Standardized survey of 70 students. Questions 1 to 5 scale
CAUSALITY
Conditions
→ A occurs before B (timing is important: cause takes place before the effect)
→ there is an association between A and B
→ there is no other variable that can lead to B
Causality is usually established based on deductive reasoning. It cannot be determined
empirically with certainty
.
Answer C
, Unit of Analysis
Level at which the variable is measured
Variables in hypotheses have the same unit of analysis
Time aspect
Cross-sectional
- Measurement at one point in time
- Relationships among variables at a point in time
- If variable A is higher than B is higher
- Basis of comparison is other respondents
Longitudinal
- Over time or time series
- Collect data from the same unit at multiple times
- If A increase in time, then B increases
Manipulation
Experimental research
→ IV is manipulated then influence on DV is measured
Ex post facto design
→ the effect of IVs on the DV is examined; IVs are not manipulated
Randomization
Randomly allocating participants to different groups (treatment and control group)
Ensures that confounding (extraneous) variables (EVs) are equally distributed across groups.
If we measure a difference in the DV between the 2 groups, it is only attributable to the IV
(treatment)
Influence of randomization on causality?
- Third condition of causality. If everything goes well then random allocation will take
care of that.
Control variables: include variables that (may) explain the dependent variable, but are not
directly relevant to our study
Monday lecture
Wednesday lecture
Thursday ISC and Tutorial
Inductive reasoning: drawing a conclusion based on an observation
• Outcome of inductive reasoning is one or more hypotheses
Deductive reasoning: conclusion follows from a set of reasons (premises)
• Begins with a hypothesis
Possibility to combine these in research process.
More valid for deductive reasoning: research process
Theory → constructs/concepts → variables
Hypothesis: expected relationship between variables based on a theory.
- Descriptive hypothesis
o Value of one variable
o 80% of students pass the QRM course
- Rational hypothesis
o Relationship between 2 variables of one case
o Students who spend more time eon the QRM course get a higher grade
Theory: a set of interrelated concepts, definitions, and propositions to explain and predict
phenomena.
→ theory is not the opposite of a fact
Construct: conceived (by the researcher) specifically for the study or for the theory building
- Difficult to observe
- Often built from concepts
- E.g., yuppies, composed of concepts (age, type of car)
Dependent variable: the variable that we want to explain
Independent variable: variable that explains (in part) the dependent variable
Moderating variable: the strength of the relationship between a dependent and independent
variable is affected by a moderating variable
Intervening variable: the influence of an independent variable on a dependent variable is
through an intervening variable.
Hallmarks of scientific research
, • Purposiveness: the supervisor or the manager of an organization has started the
research for a specific purpose.
• Rigor: carefulness, honesty, and the degree of accuracy in research investigations.
• Testability: researcher develops a hypothesis on how for instance, employee
commitment can be improved
• Replicability: we repeat the test of hypotheses again and again for the better outcomes.
• Precision and confidence: precision reflects the degree of accuracy of the results based
on the sample to what really exists in the universe. Confidence refers to the probability
that out estimations are correct.
• Objectivity: data analysis should be based on the facts of the findings based on actual
data
• Generalizability: scope of applicability of the research findings in one organizational
setting to other settings.
• Parsimony: it can be introduced with a good understanding of the problem and the
important factors that influence it.
Descriptive research
- Inventory of level of absenteeism
- How often does something occur
- Reasons/possible causes why something occurs do not attend
Causal research
- Testing cause and effect relationships
- What factors lead to difference in absenteeism?
Descriptive research → inductive reasoning → exploratory
,Causal research → deductive reasoning → hypothesis testing
Exploratory vs formal (hypothesis testing)
Exploratory
- Looser design
- Aim is to develop theories that can be tested in further research
- Ends with the formulation of hypotheses
Formal (hypothesis testing)
- Begins with the hypotheses
- Established procedure for hypothesis testing
Role of the researcher
Contrived setting: artificially created environment
Non-contrived: a natural environment
• Manipulation
• Simulation
• Controlled environment
Intensive vs extensive design
Intensive design: Intensively investigate a case
• Case study. Invite 5 students for an interview
Extensive design: Investigate many cases less deeply
• Standardized survey of 70 students. Questions 1 to 5 scale
CAUSALITY
Conditions
→ A occurs before B (timing is important: cause takes place before the effect)
→ there is an association between A and B
→ there is no other variable that can lead to B
Causality is usually established based on deductive reasoning. It cannot be determined
empirically with certainty
.
Answer C
, Unit of Analysis
Level at which the variable is measured
Variables in hypotheses have the same unit of analysis
Time aspect
Cross-sectional
- Measurement at one point in time
- Relationships among variables at a point in time
- If variable A is higher than B is higher
- Basis of comparison is other respondents
Longitudinal
- Over time or time series
- Collect data from the same unit at multiple times
- If A increase in time, then B increases
Manipulation
Experimental research
→ IV is manipulated then influence on DV is measured
Ex post facto design
→ the effect of IVs on the DV is examined; IVs are not manipulated
Randomization
Randomly allocating participants to different groups (treatment and control group)
Ensures that confounding (extraneous) variables (EVs) are equally distributed across groups.
If we measure a difference in the DV between the 2 groups, it is only attributable to the IV
(treatment)
Influence of randomization on causality?
- Third condition of causality. If everything goes well then random allocation will take
care of that.
Control variables: include variables that (may) explain the dependent variable, but are not
directly relevant to our study
