Business Research Final Exam

determining a broad problem

• Understand the problem your are seeking to research

• It doesn’t mean that something is seriously wrong

• It just could be a small issue that if it is fixed, it could improve an existing situation

Define the problem statement 

• Read literature review and previous studies 

• Ask people inside organisations that may help us to highlight specific problems 

• Set clear boundaries 

• Be specific

Preliminary information gathering

• We need to know why it is important to follow this research or to answer this research

• It helps us understand what is the problem and why does it exist and is it important

Determining the research objects

• Research objectives are reflection of research questions

• One research = One objective

• Distinguish between a question and objectives in research by highlighting or by referring to the fact that the way of writing is different

• The research question style is different than the objective 

• Research objective is the “Why” of the research

• Objectives help solve a problem or change something 

The significant of the research (importance)

• Understand what are the contribution of the research 

• theoretical and practical contribution

practical contribution

research is relevant if the problem exists in organisation or a situation that needs to be improved

theoretical contribution

•  Refers to the literature review, fostering the existing theories and literature. Contradicting results and knowledge is scattered.

3 elements of defining and refining the problem

• Relevant: Problem statement will allow you to contribute for policy for practise and also to the existing literature (practical & theoretical contribution)

• Feasible: you are able to answer the research question even with restrictions

• Interesting 

basic types of questions

exploratory

descriptive

causal

exploratory

• qualitative

• Wanting to explore the research so often the question starts with a “why” 

• We explore the question by doing stuff like interviews or surveys

• “A service provider wants to know why his customers are switching to other service providers?”

descriptive

• qualitative or quantitative

• Enables the researcher to describe the characteristics of the variables of interest in a situation

• Describe specific things or the situation

• “What is the profile of the individuals who have loan payments outstanding for 6 months and more?”

causal

• quantitative

• Cause and effect relationship between different variables

• Describe one or more factors that are causing a problem

• “Will the sales of product X increase if we increase the advertising budget?”

steps to define and refine the problem

• determine a broad problem

• define the problem statement

• determine the research objectives

• the significance/importance of the research

variable

• Any concept or construct that varies or changes in value (change occurs at various times or different people)

types of variable

• dependent variable (outcome)

• independent variable (impact)

• moderating variable (qualitative/quantitative)

• mediating variable (intervening variable)

dependent variable

• Outcome or Effect: The dependent variable is what you measure in an experiment. It's the result that you think will change when you manipulate another variable.

independent variable

• Cause or Influence: The independent variable is what you change or control in an experiment to see if it affects the dependent variable.

Moderating variable (Qualitative/Quantitative):

• Impacts the relation between independent and dependent variable

Mediating variable (intervening variable):

• Comes between the independent and the dependent variables 

hypothesis

• A proposition that we seek to examine.

• “Examining the relationship between X and Y”

• A good hypothesis must be testable, acceptable for its purpose, better than it’s opponent

hypothesis statements

• Proposition: “employee will be more satisfied when they are more motivated”

• If-then statement: “If employees are more motivated, then they will be satisfied

directional hypothesis

• The hypothesis indicates the direction of the relationship between the variables, either to be positive or negative. 

• “There is a positive relationship between x and y”

• It tells you which way the outcome is expected to go.

non-directional

• No clue about the direction of the relationship

• “There is a relation between x and y”

• Why do we use this? Relationship have never been explored, conflicting findings

• It says there will be a change, but doesn’t say if it will go up or down, more or less, better or worse.

null hypothesis

• Proposing there is no relationship between the variables 

• Set up to be rejected in order to support the alternative hypothesis

alternative hypothesis

• There is a relationship between X and Y 

• Set up to be accepted

Attitudinal scales

• Rating scales to rate an object using several responses categories 

• Ranking scales to make comparison between objects

likert scale

• The scale that is used inside the questionnaire in order to allow participants to give their answers. 

• “Scale 1 to 10 if you agree with a statement” 

• After developing on or more questions, you adopt a scale to assign numbers to the attributes of the objects, you classify the objects

four types of scales

• nominal

• ordinal

• interval

• ratio

nominal scale

• Difference: Yes (the two options are different from each other, Like if we ask about their gender, could be male/female)

• Order: No (it is not important to know which gender comes on top of which)

• Distance: No (distance between the options, we have clear options so the distance isn't important)

• Unique Origin: No (no 0 value)

• Definition: Categorizes data without any order or numerical value.

ordinal scale

• Difference: Yes 

• Order: Yes (order matters, and difference is important, because diploma is less than the bachelor)

• Distance: No (the distance doesn't matter)

• Unique Origin: No

• Definition: Categorizes data with a specific order or ranking.

interval scale

• Difference: Yes 

• Order: Yes  

• Distance: Yes (like asking about someone's happiness, could be happy or not or a little happy) 

• Unique Origin: No

• Definition: Measures data with equal intervals between values but no true zero point.

ratio scale

• Difference: Yes

• Order: Yes

• Distance: Yes

• Unique Origin: Yes

• Definition: Measures data with equal intervals and a true zero point, allowing for meaningful ratios.

validity

Make sure that the questionnaire measures what I intend to measure, or examines what I want to examine.

types of validity

• Face validity 

• Content validity 

• Internal validity 

• External validity 

• Construct validity (needs statistical analysis)

  • Convergent and discriminant validity

reliability

• Internal consistency between the questions used to measure one variable 

• Questions about the topic are related

sampling

• Process of selecting a sufficient number of elements from the population, so that results from analysing the sample are generalizable to the population.

population

• Entire group of people

• Every person working in a specific area

element

• Single member of the population

sample

• The specific members selected that are to be covered in the research

subject

• Every member of the sample is called a subject 

sampling techniques

• probability sampling

• non-probability sampling

probability sampling

• The chance of selecting each participant in the population is KNOWN and not zero and equal

• has some types of sampling which are

  • simple random sampling

  • systematic sampling

  • stratified random sampling

simple random sampling

• Each element has a known and equal chance of being selected. Randomly select participants

• Easy to understand - highly generalizable

systematic sampling

• Assign a number to every participant in your population

• Decide how large your sample size should be 

• Determine the sample interval 

stratified random sampling

• Population that includes different categories/groups

• Like population includes males/females

• two processes for stratifying

  • proportional

  • disproportional

proportional stratifying processor

• We know the number of each category/stratum (10k students, 8k men, 2k females)

• Forced to choose a certain percentage from every category/stratum

disproportional stratifying processor

• We don't know the number of each category/stratum 

• Forced to pick an equal number from every stratum

• Will divide the sample size by the number of category

non-probability sampling

• Elements don't have a known or equal chance of being selected as subjects 

• types of non-probability sampling

  • convenience

  • purposive

  • quota

  • snowball

convenience sampling

distribute the questionnaire for anyone

purposive sampling

target specific people

quota sampling

different groups, no sample frame

snowball sampling

first participant helps bring more participants

how to know which sampling to follow?

• Depending on the sample frame, means you have a list of all elements in the population. (having the list of names of an organisation) 

validity and reliability

• Face validity (no statistical inference)

• Content validity (no statistical inference)

• Construct validity (needs statistical inference) 

• Internal Validity (no statistical inference)

• External validity (no statistical inferenc

reliability tests

• Internal consistency between the items used to measure one variable 

• Measure based on the correlations between different items on the same tes

Common guidelines for evaluating (cronbach’s alpha)

• 00 to .69 = poor

• 70 to .79 = good

• 80 to .89 = excellent/strong

• 90 to .99 = Using the same question many times so bad 

• (don't repeat the same question many times, we need to have different questions to measure one variable, but these questions are related = reliability)

examining the relationship (two types) hypothesis

• correlation

• regression

correlation

• know the relationship between variables

• Indicates direction, strength, and significant of relationships among variables 

• Do a test that will allow us to understand if there is a correlation between two variables without knowing which one causes the other. 

• Not enough to examine the hypothesis 

• Rang from -1 to +1 

regression

• Examine the hypothesis

• Coefficient of determination 

• from 0% to 100%

• We know which variable causes the other

the P value (in regression)

• To accept hypothesis, the P value has to be less than 0.05

• Why? Has something to do with Margin of error and level of confidence

two types of regression

• simple

• multiple

simple regression

• Used when I have one independent variable is hypothesised to affect one dependent

multiple regression

• Used when I have more one independent variable to explain the variance in the dependent variable 

qualitative data analysis

• record the interviews

• write the transcript

• analyse the transcript

qualitative sample size

• Keep sampling until you reach data saturation

• Data saturation

  • At a certain stage of doing the interviews, you are not receiving new ideas

  • Participants are repeating the same answers 

data reduction types (how to analyze)

• coding

• categorization

coding

• Take all the words or ideas you have collected and organise them in a way that it makes sense 

• Go over the interview, identify common topics or issues and group related responses together

categorization

• Putting them into labels 

• Organise them or classify them so you can easily see how they relate to each other

• Putting each response under a specific category for example

braun and clarke (2006)

1. Become familiar with the data 

   1. Read through the data carefully and understand it well

2. Generate initial codes

   1. Identify and label important features of the data

3. Search for themes

   1. Group the codes into themes that capture a significant pattern

4. Review themes

   1. Check if the themes work well with the data 

5. Define themes

   1. Clearly define and name each theme

6. Write up

   1. Write a detailed report explaining the themes and supporting them with data