ch2 Flashcards
What makes a good business problem?
- Feasible
* Relevant
Feasibility (problem statement)
- Is the problem demarcated?
- Can the problem be expressed in variables?
- Are you able to gather the required data?
Managerial Relevance (problem statement)
•Managerial relevance (who benefits from problem solved?)
•Managers
of one company, of one industry, of multiple industries
•End users
•Public policy makers (government etc)
Academic Relevance (problem statement)
•Completely new topic •New context •Integrate scattered research •Reconcile contradictory research Academic relevance: has the problem not already been solved in prior research?
What makes a good problem statement?
Formulated in terms of: •Variables and relations •Open-ended questions •Stated clearly/unambiguously •Is managerially and/or academically relevant
What makes good research questions?
•Should collectively address the problem statement
•First theoretical, then practical research questions
-in the same order as they will be addressed in research report
•stated clearly/unambiguously
Types of theoretical research questions
- Context questions (eg “What is…”) - only if context needs elaboration
- Conceptualization questions (eg “What is…”) - only for key variables that need elaboration
- Relationship questions (eg “Which variables…” ; “How does … affect …”) - all relationships in the problem statement should be covered
Types of practical research questions
- Relationship questions
- Implication questions
Practical questions focus on specific case/company whereas theoretical questions are general and not company specific
Good to know
- The problem statement is the general question you try to answer in your research
- Research questions help to answer the ‘overarching’ problem statement, step by step
- Drawing up a problem statement and research question is difficult and time consuming, but oh-so-important. They determine the contents and the structure of your research report
What makes a good variable definition
•Informative variable name (keep it short if possible)
•Variable definition without jargon
-based on a careful literature review
-Unless very obvious (e.g.g sales, profits)
-Pitfall: examples do not substitute for a definition
•One or two supporting references per variable definition
What if many different variable definitions exist in the literature?
- Acknowledge the major differences
* End with a definition that focuses on the shared meaning across definitions or pick one definition and justify why
Independent variable
Influences the dependent variable
-in a positive way, or
-in a negative way
AKA: predictor variable, IV, denoted as X
Dependent variable
The variable of primary interest
AKA: criterion variable, DV, denoted as Y
Mediating variable
Explains why or how the IV, X, has an influence on the DV, Y. It explains the mechanisms at work between X and Y.
AKA: mediator, intervening variable
e.g.
•CEO communication style -> Employee morale -> Employee productivity
•Advertising spending -> Consumer attitudes -> purchase intentions
•R&D spending -> #of new product introduction -> Sales
Moderating variable
A variable that alters the strength and sometimes even the direction (positive negative) of the relationship between x and y.
AKA: moderator, interaction variable
e.g. X = advertising spending, Y= Sales. Mod = Recession (yes/no), OR TV advertising vs internet advertising
Theoretical framework consists of:
- Variable definitions
- Conceptual model
- Hypotheses
What makes a good hypothesis?
- Testable (measurable variables)
- Derived from theory
- unambiguously phrased
Directional vs unidirectional hypotheses
Directional hypotheses indicate the direction of the effect, whether it is positive or negative
Undirectional hypotheses indicate that there is an effect, but do not indicate a direction (should be used when outcome is even)
Null hypothesis - H0
- Expresses no relationship between variables
- Set up in order to be rejected (in favour of the alternate hypothesis)
! Null hypotheses are NOT presented in research reports
Alternate hypothesis - H1
- Expresses a relationship between variables
* = research hypothesis
Quasi moderation
MOD moderators the relationship between X and Y, but it also has a direct effect on Y. This direct effect is included in the study
Pure moderation
MOD moderates the relationship between X and Y, but it has no direct effect on Y
Full mediation
X only has an effect on Y through mediation.
Eg: to what extent are actual delays related to customers’ satisfaction with CAP through perceived delays? perceived delays = mediator.
Partial mediation
X has only an indirect effect on Y through MED, but also a direct effect on Y.
Eg: There is a direct relationship between age (IV) and blood pressure (DV), and this relationship is also mediated by weight. (partial) mediator = weight.