module 3

Question 1

providing a literature review

Answer 1

Each study builds on previous work. The purpose of a literature review is to provide a condensed overview of the key studies on a particular topic.

Are you familiar with the phrase “standing on the shoulders of giants”

The “standing on the shoulder of giants” metaphor is often used to acknowledge the work of others when undertaking research and, in particular, stresses the importance of literature reviews in scientific inquiry. That is also why google scholars homepage uses these words

Question 2

The conceptual model

Answer 2

Whereas a literature review summarizes the existing research on a topic, a conceptual model visually summarizes the new study.

Specifically, the conceptual model explains which variables are included in a study (and justifies why these are included) and how they relate to each other

Question 3

four types of variables

Answer 3

Dependent variables
Independent variables
Mediating variables
Moderating variables

Question 4

Independent variable

Answer 4

Influences the dependent variable
- in a positive or negative way

Question 5

Dependent variable

Answer 5

The variable of primary interest

Question 6

Mediating variable

Answer 6

A variable that explains the mechanisms at work between X and Y

X –> MED –> Y

CEO communication style –> employee morale –> employee productivity

Question 7

Moderating variable

Answer 7

A variable that alters the strength and sometimes even the direction (positive /negative ) of the relationship between X and Y

Question 8

Moderating variable example

Answer 8

X – MOD –> Y

Advertising spending – recession (yes/no) –> sales

Question 9

Control variable

Answer 9

anything that is held constant or limited in a research study.

Question 10

Can a moderator variable have a direct effect on the direct variable

Answer 10

Yes

Question 11

A fitting problem statement for a conceptual model where the moderator variable also has a direct effect on the Direct variable would be:

Answer 11

To what extent are X and MOD related to Y and how does the relationship between X and Y depend on MOD

Question 12

Literature review

Answer 12

The key to writing an interesting literature review is structure. A good literature review is structured around relevant themes about your subject that help to highlight similarities between prior findings.

In a good literature review you synthesize existing studies rather than summarize them. That means that you should not just simply list all the articles findings about your subject, but you should integrate previous studies and create your own story

Question 13

Never describe studies conologically

Answer 13

the key to writing an interesting literature review is structure. Describing studies chronologically is not a good way as it tends to be enumerative (boring). It is more meaningful to structure a literature review around relevant themes about your subject that highlight relationships between studies, as well as controversies and/or gaps

Question 14

Never argue that a lit review is not needed because the topic has never been researched before

Answer 14

If the topic you want to study has never been examined in prior research, you will want to find inspiration in studies on similar topics. Then draw on your reasoning powers to identify the most important common elements

Question 15

Before you visually show the conceptual model:

Answer 15

Briefly discuss the general thrust. In doing so, ensure you include a formal definition of each variable. As you should always try to build on prior research, a variable definition should be based on the literature.

Sometimes, you may come across different definitions in the literature. In such instances, we advise you to first acknowledge the major differences between the various definitions. Subsequently, you can either focus on the shared meaning across definitions. Alternatively, you can pick one definition – provided it comes from a reputable source – and justify why you will use this definition (and not the others) in your research.

Question 16

Never define your vairables one after the other

Answer 16

Instead integrate the definitions in the text that briefly describes your conceptual model

Question 17

Never use synonyms for your variable names

Answer 17

You are not writing a novel but a scientific report. Using exactly the same variable names throughout your report provides clarity

Question 18

Never define a variable by copy-pasting the first definition you encounter in the literature

Answer 18

Define your variables carefully. Make an overview of the various definitions that are offered in the literature and proceed with the shared meaning. Alternatively, justify why a specific definition fits your study best

Question 19

A variable cannot be defined by using examples

Answer 19

You first need to provide a formal definition, after which you can supplement that definition with examples. Examples can never replace a reference

Question 20

Never strive for complexity

Answer 20

A conceptual model must be parsimonious. It must be simple enough to be readily applied. If it is very complex, it becomes difficult to derive explicit predictions about real-world events from it

For example, a model with one independent variable and multiple moderators is a parsimonious model. In a similar vein, a model with multiple independent variables and one moderator is a parsimonious model.

Question 21

What is a research hypothesis

Answer 21

A tentative statement about the coherence between two or more variables

Question 22

A research hypothesis is a tentative statement

Answer 22

This means that a research study will test, using data, whether this statement is sound

Question 23

A research hypothesis pertains two or more variables

Answer 23

A main effect hypothesis is about the relationship between two variables. A mediator and a moderator hypothesis are about the relationship between three variables

Question 24

Directional hypotheses

Answer 24

Indicate the expected direction of the relationship; is the expected association positive or negative

For example: higher workloads are associated with lowe employee morale is a directional hypothesis. The hypothesized main effect is negative

Question 25

Directional hypotheses

Answer 25

Indicate the expected direction of the relationship; is the expected association positive or negative

For example: higher workloads are associated with lowe employee morale is a directional hypothesis. The hypothesized main effect is negative

Question 26

Undirectional hypotheses

Answer 26

Expect a relationship, but they do not indicate the direction

For example: Workloads are associated with employee morale is an undirectional hypothesis. The hypothesis does not indicate whether the main effect is likely to be positive or negative

Question 27

When to use undirectional hypotheses

Answer 27

Use sparingly; only use them when the theory points in two directions that are equally likely

Question 28

Why start with research hypothesis

Answer 28

Whether a study starts with research hypotheses or not, it will of course produce the same empirical findings. However, without research hypotheses, these empirical findings could be a mere coincidence.

Question 29

What makes a good research hypothesis

Answer 29

Should be testable; means that it should be phrased in terms of (measurable) variables

SHould not be based on your gut feeling but should be justified using logical arguments based on prior research studies

Question 30

Finding support for a hypothesis

Answer 30

Hypotheses are neither true nor false in an absolute sense. You can therefore never claim that you have proven a hypothesis. The word prove is not used in business science. A single confirming finding can never prove a hypothesis.

Instead of saying “proven,” we say that a hypothesis is supported by (consistent with) or not supported by (not consistent with) the data. A hypothesis is not right; it is simply not proven wrong.

When you see the word prove in a headline in the popular press, be skeptical! No single study can prove a hypothesis once and for all. A more accurate headline would be: “Study supports the hypothesis that ….”

Question 31

Choosing variable names

Answer 31

Before formulating your research hypotheses, you must decide on the names you will use to refer to your variables. Variable names hsould (i) not overpromise, (ii) leave no room for ambiguity, nad (iii) be short

Question 32

Variable names should not overpromise

Answer 32

In choosing a variable name, you should make sure it is to the point and does not overpromise

Question 33

Variable names should be unambiguous

Answer 33

In choosing a varaible name, you should make sure it is unambiguous. Avoid variable name that can be interpreted in multiple ways

Question 34

Variable names should be short

Answer 34

At the same time, you should try to make your variable names as short as possible, in the interest of readability. As you will be referring to your variables repeatedly in your research report, very long names will become a nuisance

Question 35

Guidelines to formulate research hypotheses

Answer 35

A research hypothesis proposes a relationship between two (or more) variables. The correct formulation of a hypothesis differs depending on the type of variables involved. An important distinnction is whether a variable is quantitative or categorical

Question 36

Quantitative variable

Answer 36

Captures a quantity.
E.g. household size or number of persons in a household)

Question 37

Categorical variable

Answer 37

Has different levels or categories that are not ordered along an underlying dimension

Question 38

A main effect hypothesis when botht he DV and IV are quantitative

Answer 38

If two quantative variables are related, you can expect the relationship/association between these two variables to be either positive or negative

When you expect that an increase in X corresponds to an increase in Y, the relationship between X and Y is positive. Visually, a positive relationship looks like the cloud of data points climbs to the upper right-hand corner of the graph.

Question 39

Hypotheses can be offered for variables with any number of levels, this quickly become overly complex therefore:

Answer 39

We therefore limit ourselves to the case of two levels

Question 40

How to word a mediator hypothesis

Answer 40

Expectations about a mediator effect can be formulated using two main-effect hypotheses. The first hypothesis pertains to the relationship between the independent variable X and the mediator MED, while the second hypothesis concerns the relationship between the mediator MED and the dependent variable Y. To emphasize the mediating role of MED, you can label the hypotheses as a and b components of one and the same hypothesis (e.g., H1a and H1b). As before, you need to take variable type (quantitative vs. categorical) into account.

Question 41

For three quantitative variables, the hypotheses could be formulated as follows:

Answer 41

H1a: When X increases, MED increases/decreases.
H1b: When MED increases, Y increases/decreases

Example

A research study investigated whether personalized ads are associated with more privacy concerns among consumers than non-personalized ads, and whether these privacy concerns are negatively related to consumers’ attitudes toward the advertised brand. Personalized vs. non-personalized ad is a categorical IV with two levels. Consumers’ privacy concerns is a quantitative MED, while consumers’ attitudes is a quantitative DV. The mediator hypothesis could be expressed as:

 H1a: Personalized ads are associated with more privacy concerns among consumers
          than non-personalized ads.

 H1b: When consumers' privacy concerns increase, their attitude toward the advertised
          brand decreases.

Question 42

Partial mediation

Answer 42

When the mediating variable doesnt fully affect the direct variable through the indirect variable

Question 43

How to word a moderator hypothesis
A moderator changes the strength of a relationship between two variables. A moderator can make:

Answer 43

a positive relationship stronger (more positive)
a positive relationship weaker (less positive and possibly even negative)
a negative relationship stronger (more negative)
a negative relationship weaker (less negative and possibly even positive)

Question 44

When all three involved variables (DV, IV, and MOD) are quantitative, and the MOD is expected to strengthen the positive relationship between the DV and the IV, the moderator hypothesis can be expressed as follows:

Answer 44

H: The positive relationship between X and Y strengthens when MOD
increases.

Example
A researcher expects that the number of open innovation activities firms undertake positively affects their sales. In addition, he expects that the positive effect of open innovation on firm sales becomes more positive when firms also invest more in R&D themselves. That is, he expects the effects of open innovation and R&D on firm sales to be synergistic (produce a combined effect greater than the sum of the individual effects).

The main-effect hypothesis can be expressed as:

 H: The relationship between open innovation and firm sales is positive.

The moderator hypothesis can be expressed as:

 H: The positive relationship between open innovation and firm sales strengthens as
      firms' R&D expenditures increase.

Question 45

When all three involved variables (DV, IV, and MOD are) quantitative, and the MOD is expected to weaken the positive relationship between the DV and the IV, the moderator hypothesis can be expressed as follows:

Answer 45

H: The positive relationship between X and Y weakens/decreases when MOD increases.

Example

A researcher employed by Albert Heijn expects the brand sales of mineral water to increase when price-promotion depth increases. However, he expects that the positive effect of price-promotion depth on brand sales becomes less positive for higher-priced brands.

The main-effect hypothesis can be expressed as:

 H: The relationship between price-promotion depth and brand sales is positive.

The moderator hypothesis can be expressed as:

 H: The positive relationship between price-promotion depth and brand sales weakens as
      brand price increases.

Question 46

When all three involved variables (DV, IV, and MOD) are quantitative, and the MOD is expected to strengthen the negative relationship between the DV and the IV, the moderator hypothesis can be expressed as follows:

Answer 46

H: The negative relationship between X and Y weakens when MOD increases.

Example

A researcher would like to know whether production offshoring (defined as the extent to which a firm outsources manufacturing to another country) is related to a firm’s innovativeness (measured as the number of new product introductions per year). He expects that the main effect of production offshoring on firms’ innovativeness is negative: if firms offshore a larger percentage of their manufacturing, they will become less innovative. The reason is that firms learn by doing: they learn from manufacturing products, and in the process, they may develop new product ideas. If they offshore manufacturing, they lose these sources of ideas for innovation). However, he does not expect this relationship to be equally strong for every firm. He expects that if firms invest more in R&D, they may be able to reduce the negative effect of production offshoring on their innovativeness.

The main-effect relationship can be expressed as:

 H: The relationship between production offshoring and firm innovativeness is 
      negative. 

The moderator hypothesis can be expressed as:

 H: The negative relationship between production offshoring and firm
      innovativeness weakens as firms' R&D expenditures increase.

Since the main effect is expected to be negative, it is denoted by a minus sign above the corresponding arrow in the conceptual model. Since the moderator is expected to weaken the main effect (i.e., make it less negative), it is denoted by a plus sign.

The graph on the right-hand side would support the researcher’s moderating hypothesis: for every 5% increase in production offshoring, firms’ new-product introductions hardly drop if firms spend a lot on R&D. In contrast, if firms do not spend a lot on R&D increases, the number of new product introductions drops with 4 for every 5% increase in production offshoring

Question 47

1- The IV is quantitative and the MOD is categorical (2 Levels

Answer 47

When the IV is quantitative and the MOD is categorical (with two 2 Levels), you can formulate the following hypothesis:

H: The relationship between X and Y is stronger/larger (weaker/smaller) for
Level_1_of_MOD than for Level_2_of_MOD.

Example

A researcher hypothesizes that the number of open innovation activities firms undertake positively affects firm profitability, but does not expect this effect to be uniform across all firms. Specifically, he expects that the relationship between open innovation and performance is more positive for publicly listed firms than for private firms. The moderation hypothesis can be expressed as follows:

 H: The positive relationship between open innovation and firm profitability is larger for           publicly listed firms than for private firms.

Question 48

2- The IV is categorical (2 levels) and the MOD is quantitative

Answer 48

When the IV is categorical (with two 2 Levels) and the MOD is quantitative, you formulate the following hypothesis:

 H: The difference (or: gap) in Y between Level_1_of_IV and Level_2_of_IV is
      larger/smaller (or: increases)/decreases when MOD increases.

Example

A researcher hypothesizes that men earn more than women. He expects the salary difference between men and women becomes larger for people with higher education levels. The moderation hypothesis can then be expressed as follows:

 H: The salary difference between men and women is smaller when their
 education level increases.

Question 49

3- The IV and MOD are both categorical (2 levels)

Answer 49

When the IV and the MOD are categorical (with two 2 Levels), you formulate the following hypothesis:

 H: The difference in Y between Level_1_of_IV and Level_2_of_IV is larger (or smaller) for   
 Level_1_of_Mod than for Level_2_of_Mod.

Example

A researcher hypothesizes that men earn more than women. He expects the salary gap between men and women to be smaller in service industries than in high-tech industries. The moderation hypothesis can then be expressed as follows:

 H: The salary difference between men and women is larger in high-tech industries than in
 service industries.

Question 50

how to justify hypotheses

Answer 50

You justify a hypothesis by providing logical arguments, based on the existing literature. These arguments collectively make the hypothesis plausible.

After you have built up your line of reasoning, you conclude with your hypothesis, using sentences such as:

This leads to the following hypothesis:
We therefore hypothesize:
As such:
Therefore:
Thus:
after which your formal hypothesis can be presented on the next line.

Question 51

1- Do not simply claim that author X said so

Answer 51

A common mistake is to claim that a certain author has stated that your hypothesis is true, so it must be true. Don’t do this. This author might be wrong. Focus on the consensus in the literature, rather than the word of one person.

Even if the author is right, the reader still does not understand why your hypothesis makes sense. So, you should provide reasons / explain underlying mechanisms (as to why X is related to Y, or why MOD strengthens or weakens the relationship between X and Y).

Question 52

2- Do not summarize one article after the other

Answer 52

At all times, avoid summarizing one article after the other to justify a hypothesis. Instead, create your own synthesis of the literature at hand.