Research Process & Basic Concepts of Statistics Flashcards
Key steps in the research process
- Define the research problem, the target phenomenon, what is know, what is gap.
- Develop a research plan.
- Theoretical framework
- Collect data.
- Analyze data.
- Report findings
Types of Variables:
Nominal Variables
Ordinal Variables
Interval Variables
Ratio Variables
Nominal Variables
Categories without a specific order (e.g., gender, nationality).
- Distinct categories
- No overlapping categories
- Similar/different
- No ranking is implied
- Two categories (e.g. gender) and multiple categories (e.g. nationality)
Ordinal Variables
Categories with a defined order, but no consistent difference between categories (e.g., satisfaction level: low, medium, high).
- Ranking is implied (either natural or defined)
- Does not account for the amount of differences between the categories
- Order of goodness, opinion poll …
- E.g. military ranks, fully agree ↔ fully disagree
- Note! For example opinions measured on a Likert-scale is an exception, because
although the variable is measured on ordinal scale, calculating means is allowed
Interval Variables
Numeric scales with equal intervals but no true zero (e.g., temperature in Celsius).
- Numeric variables
- Ordered scale in which the difference between measurements is a meaningful quantity
- Variables have measurement units
- Additions and subtractions are allowed
- Does not involve a true zero point where the characteristic ”disappears”
- E.g. temperature in °C or F or year of birth or pH-value (0-14)
Ratio Variables
Numeric scales with equal intervals and a true zero (e.g., height, weight, income).
- Ordered scale
- Involves a true zero point
- Can be multiplied by a constant
- E.g. length, weight, age
Identify variables in real-world situations:
Example 1: Age of participants in a survey
Ratio Variable (It has a true zero and the intervals between values are meaningful).
Identify variables in real-world situations: Example 2: Customer satisfaction levels: “Very Dissatisfied,” “Neutral,” “Very Satisfied.”
Answer: Ordinal Variable (There is a clear order, but the differences between levels are not measurable).
Identify variables in real-world situations: Example 3: Types of cuisine (Italian, Chinese, Mexican) ordered in a restaurant
Answer: Nominal Variable (There is no specific order in the categories).
Identify variables in real-world situations: Temperature measured in degrees Fahrenheit.
Answer: Interval Variable (It has equal intervals, but no true zero).
Population is
consists of all the items or individuals about which you want to draw a conclusion (e.g. personnel, products, habitants …)
Sample is
the portion of the population selected for analysis
Measuring
means linking numerical values to research objects
Observation unit i.e. statistical unit is
a single research object (e.g. person, product, habitant …)
Observation is
the measured result (value) that is related to one research object
A variable is
a characteristic of an item or
an individual. In empirical research variables are measured
Variables are classified
as either being quantitative i.e. numerical (such as a person’s
weight) or qualitative i.e. categorical (such as a person’s sex).
Numerical variables are further classified as
having either discrete or continuous values. Continuous variables are measured and can have any value within a range, while discrete variables are counted and can only take distinct, whole number values.
Continuous variables
Continuous variables can take an infinite number of values within a given range. These variables are measurable and can include any value, including decimals and fractions.
Examples:
Height (e.g., 170.5 cm)
Weight (e.g., 68.2 kg)
Temperature (e.g., 36.7°C)
Characteristics:
Can take any value within a certain range.
Often result from measurements.
Between any two values, there can always be another possible value (e.g., between 2 and 3, there is 2.5).
Discrete variables
Discrete variables have numerical values that arise from a counting process or just
separate different values from each other. (e.g. number of things) and can only take specific, distinct values, usually integers. These values are countable and have a clear distinction between one value and the next.
Examples:
Number of children in a family (e.g., 2, 3, 4)
Number of cars owned (e.g., 1, 2, 3)
Number of students in a class (e.g., 25, 26, 27)
Characteristics:
Can only take specific values, no fractions or decimals.
Often result from counting.
There are gaps between the possible values (e.g., you can have 2 children or 3 children, but not 2.5 children).
Delimitations
- Exact delimitations should be defined for the
study (target phenomenon is usually broad) - In delimitations you define what is included in the
research and what is left outside. - Theoretical delimitations: you explain which
theoretical concepts / theoretical models you will
use in your study. - Empirical delimitations: what is included in
empirical part; a specific case‐company, or
specific customer segment, or specific market. - Everything needs to be justified.
It’s crucial to narrow down the research topic by establishing theoretical and empirical delimitations to focus on specific aspects of the phenomenon.
Example from the Lecture: The lecturer provided an example where a supermarket might focus only on student customers, excluding other types of customers as part of the empirical delimitation.
Variables in figure
Qualitative
- Nominal (age, nationality)
Quantitative
- Descrete (number of kids)
- Ordinal scale (amount of students, military ranks)
- Continuous (weight)
- Ordinal scale (pain level on a scale from 0 to 10)
- Interval Scale (numerical, no zero: Tempetrature)
- Ratio Scale (numerical, with zero: Age)
which is the measurement level: age
Age:
Measurement Level: Ratio
Explanation: Age is measured on a numeric scale with a true zero (0 years means no age). It allows for meaningful comparisons and calculations (e.g., someone who is 20 years old is twice as old as someone who is 10).
which is
the measurement level: amount of students
Amount of Students:
Measurement Level: Ratio
Explanation: The number of students is a countable quantity with a true zero (0 students means there are no students). It allows for operations such as addition and multiplication, making it a ratio variable.
which is the measurement level: nationality
Nationality:
Measurement Level: Nominal
Explanation: Nationality categorizes individuals into distinct groups (e.g., American, French, Japanese) without any inherent order or ranking. The values are simply labels representing different categories.
which is the measurement level: temperature
Temperature:
Measurement Level: Interval
Explanation: Temperature is measured on a scale (e.g., Celsius or Fahrenheit) that has equal intervals, but it does not have a true zero point in the same way that ratio variables do. For example, 0°C does not mean “no temperature.”
which is the measurement level: military ranks
Military Ranks:
Measurement Level: Ordinal
Explanation: Military ranks indicate a specific order or hierarchy (e.g., Private, Corporal, Sergeant) but do not have consistent intervals between them.
which is the measurement level: gender
Gender:
Measurement Level: Nominal
Explanation: Gender categorizes individuals into distinct groups (e.g., Male, Female, Non-binary) without any inherent order.
which is the measurement level:
Department
1 Business Administration
2 International Business
3 Other
Department (1 Business Administration, 2 International Business, 3 Other):
Measurement Level: Ordinal
Explanation: Although the categories represent different departments, they are numbered and imply a ranking or order, but the differences between categories may not be meaningful.
which is the measurement level:
Year of studies ___ year
Year of studies ___ year:
Measurement Level: Ratio
Explanation: The year of study (e.g., 1st year, 2nd year) can be treated as a numeric variable with a true zero (indicating no years completed) and allows for meaningful comparisons.
which is the measurement level:
Estimate, how much you usually
Participate in lessons? ___ % of lessons
Estimate, how much you usually participate in lessons? ___ % of lessons:
Measurement Level: Ratio
Explanation: Participation percentage is a continuous variable with a true zero (0% participation means no participation) and allows for meaningful calculations.
which is the measurement level:
What is your opinion about the lecture times?
1 They suit me well
2 They suit me poorly
3 They are totally unsuitable
What is your opinion about the lecture times?
Measurement Level: Ordinal
Explanation: The responses (1, 2, 3) indicate an ordered preference regarding the suitability of lecture times but do not have equal intervals between the options.
which is the measurement level:
Have you passed the following courses? Grade
1 Consumer Behaviour 0 No 1 Yes ___
2 Marketing Research 0 No 1 Yes ___
3 Principles of Marketing 0 No 1 Yes ___
Have you passed the following courses? Grade (1 Yes, 0 No):
Measurement Level: Nominal
Explanation: The responses categorize individuals based on whether they have passed the courses, with no inherent order between categories.
which is the measurement level:
What was your average grade in high school? _____
What was your average grade in high school? _____:
Measurement Level: Ratio
Explanation: Average grade is typically measured on a numeric scale (e.g., 0 to 100), with a true zero point (0 indicates no grade) and allows for meaningful calculations and comparisons.
which is the measurement level:
Evaluate the success of your studies
1 Very good
2 Better than average
3 Average
4 Worse than average
5 Poor
The measurement level for the scale “Evaluate the success of your studies” with responses:
Very good
Better than average
Average
Worse than average
Poor
is ordinal.
Explanation:
Ordinal level: The responses are ordered from “Very good” to “Poor,” indicating a ranking of success in studies.
The differences between the levels are not necessarily equal. For example, the gap between “Very good” and “Better than average” may not represent the same degree of difference as between “Worse than average” and “Poor.”
Thus, this scale is classified as ordinal because it reflects ordered categories without consistent intervals.
which is the measurement level:
Do you wear glasses/contact lenses?
0 No 1 Yes Strength? ___
The measurement level for the question “Do you wear glasses/contact lenses? 0 No 1 Yes Strength? ___” consists of two parts:
Do you wear glasses/contact lenses?
This part is measured at the nominal level.
Explanation: The responses (0 for “No” and 1 for “Yes”) categorize individuals into two distinct groups without any inherent order.
Strength? ___
This part, where a numeric value can be filled in (e.g., prescription strength), is measured at the ratio level.
Explanation: The strength of the glasses or contact lenses can take any non-negative numeric value, has a true zero (indicating no strength), and allows for meaningful comparisons (e.g., a strength of 2 is greater than a strength of 1).
Summary:
“Do you wear glasses/contact lenses?”: Nominal
“Strength?”: Ratio
which is the measurement level:
How many hours a week do you attend to the lessons? ___ hours
The measurement level for the question “How many hours a week do you attend to the lessons? ___ hours” is ratio.
Explanation:
Ratio level: This variable has a true zero point (0 hours means no attendance), which allows for meaningful comparisons (e.g., someone attending 4 hours is twice as much as someone attending 2 hours).
The intervals are consistent; the difference between any two values (e.g., 2 hours and 4 hours) is measurable and meaningful.
Thus, this question is considered to be at the ratio level of measurement.
which is the measurement level:
How often do you practice the following hobbies?
Daily = 3
Couple of
times a
week = 2
Couple of
times a
month = 1
Not at all = 0
The measurement level for this scale (e.g., Daily = 3, Couple of times a week = 2, Couple of times a month = 1, Not at all = 0) is ordinal.
Explanation:
The responses are ranked in a specific order (from “Not at all” to “Daily”).
However, the intervals between the options may not be equal. For example, the difference between “Daily” and “Couple of times a week” is not necessarily the same as between “Couple of times a week” and “Couple of times a month.”
Thus, this scale is ordinal because it indicates an order of frequency but doesn’t guarantee equal spacing between the options.
which is the measurement level:
What is your opinion about the following statements?
Fully
agree
Almost
agree
Hard to
say
Almost disagree
Fully disagree
The measurement level for the scale described in the question (e.g., Fully agree, Almost agree, Hard to say, Almost disagree, Fully disagree) is ordinal.
Explanation:
The responses have a clear order or ranking (from “Fully agree” to “Fully disagree”).
However, the differences between the levels are not necessarily equal or measurable (i.e., the gap between “Fully agree” and “Almost agree” may not be the same as between “Almost disagree” and “Fully disagree”).
Thus, this type of data is ordinal, as it represents ordered categories but without consistent intervals between them.
Explanation:
Ordinal level: The categories (e.g., “Fully agree” to “Fully disagree”) are ordered, and the numbers represent this order.
Why not interval or ratio?: The difference between “Fully agree” (1) and “Almost agree” (2) may not be the same as the difference between “Almost disagree” (4) and “Fully disagree” (5). The values indicate order but not exact distances between the responses.
Thus, this scale is ordinal despite the numerical coding.
Compressive and descriptive methods
- Statistical estimation
- Compressing data– mean etc.
- Assessing distributions
- Presenting the results in a less accurate yet more easily interpretable format
- Presenting a large quantity of data by a considerably smaller number of figures
Methods supporting conclusion making
- Testing statistical significance
- Defining probability of the results
- Concluding how probable it is that the phenomenon observed in a sample studied
applies also in the whole population the study is interested in - Notice! Statistical significance vs. practical significance