Research methods: all lectures

Flashcards of all the theory studied in the course Research Methods for Business and Economics

What is the key focus of applied research?

to improve understanding of a specific business problem

How does applied and basic research differ?

basic research: the purpose is to create or increase knowledge, build theories that will serve as a foundation of further studies

How can researchers maintain high ethical standards?

By following the hallmarks of scientific research: rigor, purposiveness, generalisability, testability, replicability, precision and confidence, objectivity, parsimony

Why do researchers use samples?

Feasibility. It takes too much time, money, etc. to include the population in the study. In many cases it may also be impossible to actually identify every member of the population.

Why should research start from theory?

Clearly, a theory- rather than results-driven approach should generate a less contorted, more coherent set of predictions that emanate from or build upon the underlying perspective.

What is the purpose of a conceptual framework?

to identify key concepts
A conceptual framework illustrates the expected relationship between your variables. It defines the relevant objectives for your research process and maps out how they come together to draw coherent conclusions. Tip You should construct your conceptual framework before you begin collecting your data.

Explain the hypothetico-deductive approach

It is a 7 step process:
1. Identify a broad problem area
2. Define the problem statement, obejctive and research question and conceptual framework
3. Develop a testable hypothesis (2 requirements)
4. Choose measures for the variables in the hypothesis
5. Collect data
6. Analyse data
7. Interpret data

What are the 2 requirements of a hypothesis?

testable (possible to collect empirical data and produce results) and falsifiable (can be refuted through experimental observation)

What is the difference between inductive and deductive approach?

Deductive: from general to specific; start from theory to find sth
Inductive: from specific to general; develop a theory

What are the two philosophies of science and what do they study?

Ontology and Epistemology Ontology: the nature of reality (observe reality) —> objectivism, subjectivism (constructionism) Epistemology: the nature of knowledge —> positivism, interpretivism, critical realism

Explain the different types of ontology

Objectivism: every research sees reality from the same angle; there is one general truth
Subjectivism (constructivism): every researcher sees reality from different perspectives; no objective reality

Explain different types of epistemology

Positivism: based on observed facts, existing and measurable
Intepretivism (phenomenology) : specific for particular situations; the viewpoints of the researcher also count; knowledge is built through ideas and interpretation
Critical realism: there is objective knowledge but there is also room for interpretation; objective knowledge is real but cannot be easily observed, also some phenomena cannot be measured directly

Philosophies used for quantitative research

Deductive, Positivism, Objectivism

Philosophies used for qualitative research

Inductive, Intrepretivism, Subjectivism

difference between a strategic (managerial) question and a research question

A strategic questions is not the same as a research question
1. Strategic question: relevant if it relates to an existing problem in the company; provides a solution for that particular company
2. Research question: relevant if nothing (much) is known about the topic or if the knowledge is scattered and unorganizered or the results are contradictory or not general; can help a manager make better judgments; helop understand a phenomenon in more general terms

The Watson box

What are the purposes of a literature review?

1. To report on scientific literature that forms the base of my study
2. To structure exisiting studies
3. To avoid reinventing the wheel
4. To gather useful insights into the research topic
5. To introduce relevant key terms, concepts, variables, definitions and relations
6. To obtain useful insight of research methods used in other studies
7. To enhance testability and replicability
8. Develop my own framework
9. Relate finding to others’ work
10. To gain new perspective

What are variables?

concepts that are measurable

What is theory?

A generalisation about variables and their relationships; helps predict outcomes and make decisions; explain why variables are related or why observation happens

What is framework?

Shows how the variables are related = representation in a visual way what the research is going to study

What is hypothesis?

A provable statement relating 2 variable; a particular relationship is phrased and there is some kind of an expectation

Where does theory come from?

from observation → that’s why it can sound abstract but it actually comes from empirical analysis

What are the 3 stages of developing a theory?

1. Defining and clarifying concepts, terms and making them operational (measurable concepts are called variables)
2. Developing the conceptual framework that describes the relationships between the variables
3. Using the conceptual framework to develop theories that explain patterns and connections found in the research material

What is the difference between providing the conceptual framework in the beginning rathee than the end of the research process?

1. At the beginning: structured approach (hypothetico-deductive: generalised insights guide material and data collection); based on preliminary (meaning prior) theory, concepts, hypotheses; saves time; diminished quality of final product (why? : insights are sometimes limited and reluctance to change if needed because all the work has already been done)
2. At the end: grounded theory approach (inductive- starts from specific data in order to gather more generalisable insights); no prior theory; takes time; only clear near the end whether the output is worthwhile, but better, more novel outcome

What are concepts?

Generally accepted collection of meanings or characteristics associated with phenomena, objects, behaviours, situations

Acquired over time through experience, shared usage

Concepts go beyond dictionary definition

What are constructs?

they are more abstract and less tangible than concepts

specifically developed for research → outside of the field the word has less substantial meaning)

the more inductive approach → better to speak of constructs

What is definitions?

Used to undertand and measure concepts and avoid confusion. Very important to use not dictionary definitions (not rigorous enough) but the operational ones found in literature.

What is operational definition?

Stated in terms of specific testing and measurement criteria; should help researcher to classify the objects of study

What is variables?

Make concept measurable

Anything that can take varying values

What are the types of variables?

1. Independent = predictor, antecedent, cause; influences the DV (sleep quality)
2. Dependent = criterion, response, consequence, effect; the one the researcher wants to explain (quality of my work)
3. Moderating (MV) = moderates the relationship between IV and DV; in other words, the relationship between IV and DV depends on another variable which affects the direction and strength of the relationship (mental health status)
4. Mediating (intervening =IVV) = used between the time IV starts to influence DV and when the impact on DV is felt; helps to understand how IV affects DV by being between them (alertness)
5. Control (CV) = included to avoid bias in results

What are the conditions for causal relationship between a IV and DV? Which of the conditions in the most challenging to meet? Discuss how this challenge can be overcome.

1. IV and DV covary (both change simultaneously)
2. IV precedes DV (time sequence)
3. No other factor should be a possible cause of change in DV (if so → control variable to separate effects)
4. A theory (=logical explanation) must explain why IV affects DV

The most challenging condition to meet is 3, because it is very difficult to provide an environment where no factor can intervene with the DV. A solution to this problem can be to prepare a control variable which can separate effects and provide us with the results of just our IV influencing our DV.

What are the types of conceptual frameworks?

1. Cause-and-effect relationships = based on hypothetico-deductive approach
2. Process framework = relationships between variables are based on logic and proper order
3. Hierarchial framework = concepts are related through a higher/lower position on a scale/hierarchy
4. Maps and coordinants = locates concepts on horizontal and vertical scales of the map; 2 or more dimensions
5. Pairs of opposites = competing elements that are in a constant tension to see if my data is inlined with my theory
6. Exchange and equilibrium = exchange between supply and demand → equilibrium price and quantity
7. Similarity = groups of similar themes clustered together

What is the Ockham’s razor principle?

shave off everything that is not necessary - leave only the key concepts (law of parsimony - simplicity is key)

What is the difference between a proposition and a hypothesis?

Proposition = statement about how concepts are related but not necessarily testable or measurable

Hypothesis = statement about how concepts are related but it is testable and measurable (also falsifiable)

What are the 4 most common types of hypothesis?

1. Correlational = variables occur together without causing each other; no causality, only correlation
2. Explanatory (or causal, directional) = change in IV causes a change in DV; there is a direction os influence
3. Null = indicates that variables are not related
4. Alternative = indicates that there is a difference or relationship

What is research design?

plan for collecting, measuring and analysing data

It is about:

* quantitative/qualitative study
* exploratory, descriptive and/or causal study
* primary and/or secondary data
* time horizon
* unit of analysis
* research methods

What is the difference between quantitative and qualitative studies?

1. Quantitative: to test propositions
2. Qualitative: to explore new phenomenon and create new propositions

What is the difference between exploratory, descriptive and causal studies?

1. Exploratory
* clear idea is lacking (however we are interested in a particular phenomenon)
* new area of study → we need to define concepts → inductive approach
* results non-generalisable

* often qualitative study

2\. Descriptive

* data describes the topic of interest
* only about description, none causality, only correlation
* clearly stated research quesition

3\. Causal

* test whether or not a variable causes another variable
* the hypothetico-deductive approach

What is the difference between primary and secondary data?

1. Primary: gathered first-hand - you generate the data using interviews, surveys, observation, experiments
2. Secondary: gathered through existing sources or by others for other purposes; content analysis

What are the 4 criteria for evaluating secondary data?

1. Timeliness: don’t wait too long to gather the data
2. Accuracy: precise enough
3. Relevance: relevant to measure your concepts
4. Costs: getting access to databases can be costly

What are the types of studies based on the time horizon they occur?

1. Cross-sectional: data is gathered at one point in time (one shot observation of a phenomenon); for a lot of subjects but all at the same time) (survey of customer satisfaction since launching a new product)
2. Longitudinal: repeated across a period of time; one variable across a period of time (GDP Spain 2001-2010)
3. Combination of both: panel data (GDP growth of particular countries over a particular time period)

What is unit of analysis?

Derived from the research question

Tells us which objects are going to be studied and how the level of aggregation of the data collected (industry, company, subsector)

Unit of analysis doesn’t always correspond to the respondent (ex. study of internationalisation strategies of small companies through interviews with CEOs → unit of analysis is the company, CEO is the source of data)

What are the types of interviews?

1. Structured (=pre-coded, close-ended) → researcher leads the convo; fixed order of questions and no deviation from the list of questions; quantitative and qualitative research
2. Semi-structured (mixed) → list of questions but also a leeway to deviate from it; suitable for qualitative exploratory research
3. Unstructured (open-ended) → repondent leads the convo; suitable for qualitative exploratory research

What are the different types of interview questions?

1. Introductory: shows you have done your research and now trying to get to the specifics
2. Follow-ups: to elaborate further on a question
3. Probing question: when the first answer is unclear or incomplete
4. Direct questions: directed at the interviewee personally - be careful
5. Indirect: not direct at the interviewee personally
6. Interpreting questions: the confirm that you have understood the info correctly

What are the specific techniques for individual interviews?

1. Funneling technique: from broad, open-ended questions to narrow, more specific ones; used in unstructured and semi-structured interviews
2. critical incident approach: helps to understand respondent’s lived experience; respondent asked to choose an incident and explain in detail (suitable for semi-structured interviews)

What are the advantages of using individual interviews as a research method?

* non-verbal cues
* visual aids
* rich, in-depth data
* new insights (first-hand data)
* possibility to clarify answers and questions

What are the disadvantages of using individual interviews as a research method?

* time-consuming
* representativeness is a problem because you are creating the sample yourself so you tend to choose people who have input
* risk of bias - your info comes from a personally selected sample

What are the 2 types of group interviews?

1. expert panels → bring together experts of a particular field
2. focus groups → a panel of people all familiar with the topic but not considered experts → a good mix of different backgrounds (better to have people with divergent views)

What nature of data is suitable for focus groups?

qualitative with unstructured or semi-structured questions to achive free-flowing, yet focused discussion on a clearly-defined theme

What are the advantages of focus groups?

* in-depth digging into questions
* different views
* cost and time efficent
* serendipidy: you expect a result but end up with sth completely different or new
* snowballing effect: one insight leads to another insight → new and better completed insights
* exploratory for further research

What are the disadvantages of focus groups?

* individuals may dominate the discussion
* reluctance to speak up
* group think (risk of socially desirable answers)
* the data may not be truly representative

→ use with combination with other research methods!!!

Explain the delphi-technique. What is it, when is it used and why?

Suitable for focus groups.

What?: a structured way of brainstorming; you give anoymous questionnaires and get the responses summarized to minimize group think. Delve deeper into the gathered insights.

Why? To avoid group think and dominant individuals but also come up with a consensus.

What is coding and what does it entail?

= drawing out usable material from the data collected; assigning categories to data after transcribing

What is observation?

A technique to collect data on actions and behaviour in the field

In a natural environment or lab setting.

What are the key question when designing an observation?

1. Who are you going to observe?
2. When are you going to observe them?
3. What are you going to observe? (behaviour, relationships, habits, movements)
4. How? (how many people and with what tools - cameras, etc)
5. Where as a location? : artifical or natural setting

When are observations considered scientifically sustained?

* when it answers a specific research question
* proves a reliable and valid account of what happened
* observations are consistent and lead to generalisability
* the observation help measure what you want to measure to obtain results

What are the 3 big groups of observations?

1. Participant vs Non-participant
2. Structured vs Non-structured
3. Concealed vs Unconcealed

What is the difference between participant and non-participant observation?

* Participant: the observer is a member of the observed group and the ones observed know that; popular with constructisits; used in case studies and grounded theory studies
* Non-participant: the observer is not a member of the observed group and not involved in the actions of the actors, he observes them from a distance
* Ethnography: tries to explain people’s behaviour in cultures; focus on the interaction between people; used in case studies; problems with replicability and generalisability because you cannot simulate the same conditions to get the same results twice

What it the difference between structured and unstructured observation?

* Unstructured: highly opened observation; suitable for exploratory research because you don’t know what answers you are gonna get; useful for GENERATING hypotheses; no restriction on what is observed → more room for interpretation tho so a risk of bias occurs
* Semi-structured: observational check-list as a memory aid
* Structured: quantifiable and pre-coded; useful for TESTING hypothesis; making use of an observation schedule (tally chart); activity sampling = collecting data based on occurencies and frequencies

What is the difference between concealed and unconcealed observation?

* Concealed: members of the observed group are not told that they are being observed to get rid of a validity threat that they will behave differently if they know they are being observed; problem: etical drawback - recording someone without permission
* Unconcealed: members of the observed group know they are being observed; risk: may influence behaviour and lead to reactivity

What are the characteristics of good field notes?

* use exact quotes when possible
* protect confidentiality
* be precise and exact without changing anything
* be objective and include all releavant background information to reduce the threat of bias

What is a coding scheme and when is it used?

Used with observation

It is a predetermined categories for recoding what is observed

What are the advantages of observation as a research method?

* directness → no reliance on reports by others
* behavioural data is collected without asking questions
* less demanding for participants than questioning
* data collected at the time the event occurs -→ no retrospective bias
* captured in it’s natural environment
* insight into a dynamic phenomena that cannot be described with just words
* possible to observe a certain group of individuals that otherwise it would be difficult to obtain info for

What are the disadvantages of observation as a research method?

* less suitable for cause-and-effect or complex relationships
* validity problem: reactivity, behaviour of those being observed can change ovver time
* risk of observer bias
* time-consuming and expensive
* limited as a method to learn from the past (replication is not always straight-forward)
* cognitive thought processes of individuals cannot be captured

How does ontology relate to qualitative and quantitative research? Discuss.

There are 2 types of looking at reality: obejctivism and subjectivism also called constructivism. The ontologic view of qualitative research suggests that we should use a constructivist approach, whereas for the quantitative research- the objective approach is more suitable.

Briefly discuss the role of the IV in scientific research.

The independent variable precedes the dependent variable and influences it. It cannot be changed by other variables. A variable that a researcher directly manipulates (changes)

A toy manufacturer has developed a prototype of a toy for children aged 2-6 years. The prototype has entered the testing phase to check for children’s reactions as well as safety issues. To that end, the company asks a university department to set up the test.

(a) Describe how observation as a research method could enable the testing by discussing:

* the 5 key questions as part of the research process
* the way of sampling

(b) discuss one disadvantage related to the method.

(a) Who are we observing?: children aged 2-6 years

When are we observing them?: in the afternoon when they are free and they normally play with their toys

How? : How many kids?: a good sample of 50 kids (10 from each age). How are we observing them?: concealed observation - from a distance through cameras to see how they react and behave in their most natural setting to assess the safety risks; also non-participant observation; and semi-structured observation using an observational checklist so that we can have room for change based on their behaviour while still complying with certain standards from children’s toys.

What are we observing?: children’s reactions as well as safety issues

Where are we going to observe them?: in kindergarten because there we can observe children from the same age groups all at the same time while they are still in their confortable natural setting which ensures that they will act normally.

The way of sampling? :

(b) one disadvantage is that the behaviour of those being observed can change over time especially when involving kids who are extremely unpredictable.

In what type of research is experiments as a research method used?

in deductive research, for cause-and-effect studies (not in correlational studies)

What does treatment mean in experiments context?

The manipulation of the IV by creating different levels of IV to assess impact on DV. It is required to examine causal effect of an IV on a DV

What are the 2 types of experiments?

* field experiment - natural environment in which subjects normally function; limited interference of the researcher
* lab experiment - creating an artificial environment; more interference of researcher; you have your variables under control

Effects of pay on performance of employees.

What is the relationship? Suggest a hypothesis.

What is the IV? What is the DV?

What is the treatment?

Relationship: the effect on pay on performance to performance

Hypothesis: Companies which implement the pay on performance strategy have a higher performance and more productive workforce.

IV: salary

DV: performance

Treatment: have 2 departments. We pre-test both of them so that we can see if the treatment as had any results. First one raises the salary pay (that is the treatment) and the second one doesn’t (it is the control group). Then we do a post-test to see if any changes have occured. Based on our hypothesis we expect no change in the second department and increased productivity in the first department.

What is the main difference between observation and experiments as a RM?

Observation is about behaviours, talks and meeting, pros and cons of a certain decision, not about causal relationships.

Experiments focus on a certain causal relationship and what to prove/disprove it. To do this we have to control all the variables that may covary with DV.

Which variable is very important to control in an experiment?

The contaminating variable because otherwise we will never be able to assess the true effect of IV on the DV.

How do we control the contaminating/nuisance variables in an experiment?

* matching groups: deliberately spreading contaminating characteristics in all groups so that they appear equal
* randomization: randomly distribute all individuals all with equal chance so that inequalities are distributed based on the laws of random distribution

What are the disadvantages of using matching groups and randomization as techniques of controling the contamination variable in an experiment?

* matching groups: not always possible to make each group equal; also sometimes we are not aware of all contaminating factors so sample bias is present
* randomization: works only for larger samples; through normal distribution we can spread the contaminating factos thus reducing sample bias

What is the difference between external and internal validity?

* Internal validity is the confidence we place in the cause-and-effect relationship (to what extent we can be confident that IV influences DV); the higher the internal validity, the better we can claim the relationship is causal
* External validity: does the result from the research transfer to other fields, times, settings, it is generalisable?

Discuss the external and internal validity on lab experiments.

Lab experiments have high internal validity but low external one. Therefore, it is better to first do a lab experiment and then a field one so that we can ensure both types of validity.

What are the validity threats during experiments?

* history effects - unexpected event occurs which threatens the causality between the IV and the DV, thus threatening the internal validity.
* maturation effects - due to the passage of time people evolve; threat to internal validity
* testing effects: effects from exposing participants to pre-tests and post-tests; main testing effect: people knowing they are being studied will behave differently - threat to internal validity; interactive testing effect: people part of the study may pay closer attention than non-participants thus threatening external validity because it is hard to generalise on a mass scale
* selection bias effects: people chosen for a lab study may differ from people chosen irl (external validity threat); people who participate for the money - threat to internal validity
* mortality effects: people dropping out of the experiment (internal validity threat)
* statistical regression effects: caused by the selection of subjects on the basis of extreme scores or characteristics → results move towards the mean → hard to see If they moved towards the mean because of a random effect of the extremes moving towards the mean or because the treatment actually worked (internal validity threat)
* instrumentation effects: changes in observation in the measuring instrument or in the observer (internal validity threat)

What are the types of experimental design?

* Quasi-experimental design: no random assignment of subjects to groups; often no control group; often no control over treatment
* True experimental design: random assignment; treatment designed by researcher; control and treatment groups combined

What are the types of Quasi-experimental design?

* Pre-test and post-test experimental group design
* Post-test only with experimental and control groups
* Time series design

What is a pre-test and post-test experimental group design? What are it’s strenghts and weaknesses? When is it used?

No control group; self-selected sample → no randomness

When to use?: interested in the relationship between 2 variables and the making of a control group is challenging/impossible

Threats to validity: pre-test and post-test = testing effects, both main and interactive; sth may happen = history effects; takes too much time = maturation and mortality effects

What is a post-test only with experimental and control groups design? What are it’s strenghts and weaknesses? When is it used?

A Quasi-experimental design. A pre-test is not easy because the control group and the experimental group may not be the same thus introducing a contaminating factor; also lack of randomness

The testing effect is not a challenge here.

Mortality and maturation effects may pose a threat, tho.

What is a time series design? What are it’s strenghts and weaknesses? When is it used?

A Quasi-experimental design. Data of the same variable is collected at regular intervals to assess the impact of treatment over time.

Not randomized, but the time dynamic adds rigor to the research.

Maturation effect: during the passage of time a lot can happen including unexpected events thus history effects become part of the story as well

What are the 4 types of true experimental design?

* Pre-test and post-test experimental and control group design
* Solomon 4-group model
* Factorial design
* (Double) blind studies

What is pre-test and post-test experimental and control group design?

A true experimental design. Both groups are randomized and exposed to a pre-test and post-test; only the experimental group is exposed to treatment.

Validity threats: testing effects; maturation effect- you cannot keep under control all the people; mortality effect

What is the Solomon 4-group design?

A true experimental design. Increases external and internal validity. 2 groups receive a treatment (2 experimental groups), 2 do not (control groups). 2 of the groups receive a pre-test (an experimental groups and it’s control group) , 2 do not.

Validity threats: perfect but very costly; if anything happens it is the mortality effect

What is a factorial design?

A true experimental design. You test effect of 2 or more treatments at the same time on the DV. (different bus types and different fare rates simultaneuosly).

History effect → you cannot control the environment.

What is a (double) blind study?

A true experimental design designed to avoid possible bias. (placebo vs actual effect)

What are the advantages of experimental design?

* good for establishing causality between 2 variables
* manipulation of IV is easy
* contamination is controled
* a control group is used
* replication helps to find average effect of IV on the DV

What are the disadvantages of experimental design?

* artificiality of the lab setting
* generalization of non-probability samples
* number of variables is more limited compared to surveys
* studies the present and the future, not the past
* costly
* ethical limits

What is content analysis and what is its purpose?

The method of content analysis enables the researcher to analyze (large amounts of) textual information and systematically identify its properties, such as the presence of certain words, concepts, characters, themes, or sentences. To conduct a content analysis on a text, the text is coded into categories and then analyzed.

What are the advantages of content analysis?

* gives structure and numerical interpretation to text (from qualitative to quantitative data)
* participants not required
* data is immediately available
* no ethical challenges
* broad applicability

What are the disadvantages of content analysis?

* depends on input (check for authenticity, credibility and representativeness of the articles)
* coding procedures are subject to intepretation biases
* focus on measurement of variables, not on theory

T/F: Content analysis in only appropriate for deductive research.

False

T/F: Content analysis is appropriate for developing propositions

True

What is grounded theory approach?

An approach for developing a new theory → inductive approach (from specific to general)

Data collection and analysis proceed in tandem (the data shapes the theory and reshapes it over and over again till we reach theoretical saturation)

Useful for generating new concepts and ideas → exploratory research

What is theoretical sampling and when it is used?

Used in the grounded theory approach. It involves the process of data collection to generate a theory

What are the 3 types of coding in the grounded theory approach?

1. Open coding: take the data and label it
2. Axial coding: make connections between the codes and identify causal relationships (check if the relationships hold true for the new data as well)
3. Selective coding: identify core phenomena which will be at the centre of your theory

What is theoretical saturation?

Stopping rule for identifying new categories/cases (stop when nothing new is emerging anymore or new categories don’t improve the understanding of the phenomenon)

What is the difference between a concept, a category, a property, a hypothesis and a theory?

concept: collection of codes of similar content

category: group of similar concepts

property: attribute of a concept/category

hypothesis: initial idea of the relationship between concepts

theory: a set of categories forming a theoretical framework and explaining the relevant phenomenon