febuary 10th Flashcards
step-by-step approach to measuring abstract concepts (constructs) such as implicit attitudes, brand loyalty, and extroversion. L
We have to measure those concepts somehow
Construct and variable
1. Identify and develop construct
2. Define the construct
Conceptually aka constitutively: what is it?
Operationally: how do you measure it? (can use scale)
3. Develop a measurement scale
4. Test and evaluate the measurement scale
identify and develop the construct
This is the first step where you choose the concept (or construct) that you want to measure. Constructs like brand loyalty (how strongly a customer identifies with a brand) or extroversion (a personality trait) are broad, abstract ideas that need to be measured in a meaningful way.
Example: You decide you want to measure brand loyalty to understand how emotionally attached consumers are to a brand
define the construct
Conceptually (or constitutively): This is a broad definition of the concept, focusing on its essence or theoretical meaning. It’s how you describe the construct in general terms.
Example: Brand loyalty conceptually means a consumer’s emotional attachment to a brand that results in repeat purchases and commitment to buying from that brand in the future.
Operationally: This refers to how you translate the theoretical concept into something measurable—how you will measure it in a real-world context.
Example: Operationally, you might measure brand loyalty by looking at repeated purchases of a brand or how much more willing a consumer is to pay for a brand over a competitor.
You might use a scale for this: For example, a 1 to 7 scale where 1 means “not loyal at all” and 7 means “extremely loyal,” and then you use this scale to ask survey participants about their likelihood of repurchasing the brand.
Develop a Measurement Scale
Now you need to create a way to measure the construct. A measurement scale is a tool used to quantify the construct. This can involve surveys, questionnaires, or even physical measurements depending on the concept.
Example: For brand loyalty, you might develop a series of questions or statements like:
“I would recommend this brand to a friend.”
“I always buy this brand over others, even if the price is higher.”
You could then rate these statements on a Likert scale (e.g., 1 to 5, where 1 means “strongly disagree” and 5 means “strongly agree”).
Test and Evaluate the Measurement Scale
This step is about evaluating how well your measurement scale works. You test it to see if it measures what it’s supposed to measure and if it does so consistently and accurately
Conceptual definition
Description at the theoretical level
What is it
Brand loyalty: consumers emotional attachment to a brand that results in repeat purchase behavior in the present and commitment to purchase the future
Operational definition
Translation into observable phenomena
How do you measure it
Ex: brand loyalty. Looking at behaviour
Brand loyalty definition: higher willingness to pay, advocating for a brand
Brand loyalty: the degree to which a consumer: repeatedly purchases the same brand, is resistant to switching despite costs of not doing so
Brands in the middle are struggling to come up with value propositions (the gap is similar to j crew in positioning)
Lets say we wanna examine peoples extroversion
Can use the big 5 personality traits scale to determine this
Dimensions:
Openness to experience
Curiosity, flexibility, imagination, artistic sensibility
Conscientiousness
Discipline, organization, dependable
Extraversion
Outgoing, upbeat, friendly, assertive, gregarious
Agreeableness
Sympathetic, trusting, cooperative, considerate
Neuroticism
High in neuroticism means you experience more negative emotions
Implicit attitudes
People may hold unconscious attitudes and may not reveal them
IAT: implicit association test
Easier to sort 2 concepts tgt if they are closely associated in our mind
Racial category → sort male faces based on race
Expose peoples internal biases and categorizations of people
There are issues with IAT
Goal is to educate the public about hidden biases
Reliability in IAT
reliability: ability to provide consistent data
Re test: does the scale give similar results over time?
Internal consistency: do different versions or portions of the scale give similar results?
Iat : Fails with Test restest
Iat: Not internally consistent, wont get the same results each time
Validity:
measuring intended target
different validities
contructZ:
Convergent validity: is the scale measuring the same thing as similar scales
Discriminant validity: is the scale measuring something different form different scales
Content validity: does the scale include the right questions and elude the wrong questions?
Face validity: does the scale seem aporiiriate fo what it is supposed to measure
Criterion:
Concurrent validity: is the scale correlated with current related behaviours
Predictive validity: does the scale predict future behaviour
How does the iat do
Does face validity
Does content validity
Fails at predicting behavior
Types of variables
Categorical
Data that reflects groups of categories (brands of cereal, colours of the rainbow)
Numerical
Data that are numbers based (age, height, weight)
types of measures
intervak and ratio
Nominal
vDescribe categories
Ex: gender, student
Math: count, mode, frequency
When you ask someone to calculate the avg, but people think pf the numbers that come to mind frequently rather than the avg, that will bias your estimation of how much you spend in a month, we use the mode as a way to calculate the avg expenditure but we discount the extraordinary purchases we make
ordinal
Describing ordered categories
Ex: 1st vs second place, willingness to purchase
Math: count, mode, frequency, median
intervak
Describing numeric variables
Ex: temperature, willingness to pay
Math: count, mode, frequency, median, mean and standard deuviation
Like you cna have negative degrees
ratio
There is a true zero that exists: the difference is the zero means the variable does not exist (cant have neg number) ex: zero degrees celsius does not mean the absence of temperature its just cold, therefore temp would be an interval scale, whereas price would be a ratio variable
Zero: absence of that variable → cant have negative price
Describing numerical variables +true zero
Ex: price, time, etc
Math: count, mode, frequency, mean, standard deviation, % differencee
examples of nominal ordinal interval ratio
Frequency of purchase→ this would be ordinal because we treat purchase likelihood as a numerical scale, but it should be treated as an ordinal scale→ from 1-7 dont know the differences, if you select 6 (from 1-7) what dpes this mean? If someone compares 3 how would you compare them? The numbers are not meaningful bc you cant differentiate the meaning of the difference between them= ordinal
Ethnicity would be nominal
Fahrenheit→ interval
age—> ratio , cant have a negative age
Purchase likelihood→ interval
Number of purchases→ ratio
thematic coding
Probably will use for group project
Purpose: identify patterns, themes or trends with open ended responses in surveys t improve brand equity
Thematic analysis can be used for sentiment analysis, generally thematic analysis is a more powerful tool- generates more actionable insights
Provides deep, nuanced insights into the underlying feelings and motivations behind customer feedback. Constructs are defined by the customer
Sentiment analysis: looking at people’s feelings and categorizes peoples feelings towards something
Thematic analysis is powerful, more than sentiment
sentiment vs thematic
Sentiment analysis focuses on emotional tone (positive/negative/neutral), while thematic analysis explores underlying patterns and themes that explain those feelings.
Thematic analysis is more powerful because it gives you deeper insights into customer motivations and brand perceptions, which can be translated into actionable strategies for improving brand equity.
