Lecture_5 Flashcards
(19 cards)
Nominal scale
Basic level of measurement where numbers serve as tags or labels to categorize or
identify variables without any quantitative value (e.g., gender, social security number)
Numbers assigned to runners
Lowest level of
measurement
Ordinal
scale
Numbers indicate order or rank, showing relative position, but not
the magnitude of difference (e.g., product ratings - 1st, 2nd, 3rd)
Rank order of runners
Interval
scale
Equal intervals between values, but no true zero point.
Differences between values are meaningful
(e.g., temperature in Celsius or Fahrenheit)
Performance rating on a 0 to 5 scale
Ratio
scale
All characteristics of interval scale with a true zero
point, allowing for a full range of mathematical
operations. (e.g., weight in kilograms, age in years)
Time to finish in minutes
Highest level of
measurement
Scaling Techniques: Comparative Scales
Comparative Scales
* Directly compare two or more objects
* Results interpreted in relative terms (nonmetric scaling)
Examples:
* Rank order your favorite soft drinks
* Choose between Coke or Pepsi (paired
comparison)
Comparative scales reveal preferences relative to each other
Scaling Techniques: Noncomparative Scales
Noncomparative Scales
* Evaluate objects independently (metric scaling)
* Results are on interval or ratio scales
Examples:
* Rate Coke on a scale from 1 to 7
(1=not preferred, 7=greatly preferred)
* Do the same for Pepsi
Noncomparative scales give absolute preferences, suitable for statistical analysis
When and Why Use Comparative Scales
- Detect Subtleties: Ideal for small differences in well-known items (e.g., Pepsi vs. Coke)
- Simplicity: Offers an easily understood and applied method for scaling
- Fewer Assumptions: Requires less theoretical background on the constructs being measured
- Minimize Bias: Helps reduce halo or carryover effects in sequential judgments
- Decision-Making: Useful when respondents need to consider trade-offs
- Specific Contexts: No need to generalize beyond the items being compared
- Data Sufficiency: When ordinal data meets the analysis requirements
Noncomparative Scales
Continuous Rating Scale:
* Basic Characteristics: Respondents place a mark on a continuous line.
* Example: Used to measure reactions to TV commercials.
* Advantages: Easy to construct.
* Disadvantages: Scoring can be cumbersome unless computerized.
A continuous rating scale allows
respondents to place a mark along a line
representing a spectrum of responses.
Forms vary considerably.
* Reduces Central Tendency Bias: Minimizes extreme or middle-point selections common in discrete scales
* Fluid and Flexible Responses: Offers respondents a nuanced way to express opinions
Itemized Scale:
Likert Scale
Semantic Differential
Stapel Scale
- Scale Diversity: Itemized scales can vary greatly
- Scale Categories: Commonly range from 5 to 7 options.
The choice affects the level of granularity in responses - Balanced Scales: Offer a neutral option for more nuanced
feedback - Forced Choice: Force respondents to choose a category
Itemized Scale:
Likert Scale
- Basic Characteristics: Measures the degree of agreement on a scale from 1 (strongly disagree) to 5 (strongly agree).
- Example: Commonly used for measuring attitudes.
- Advantages: Easy to construct, administer, and understand.
- Disadvantages: More time-consuming compared to other rating scales.
Itemized Scale:
Semantic Differential
- Basic Characteristics: A seven-point scale with bipolar labels (e.g., good-bad, happy-sad).
- Example: Used to evaluate brand, product, and company images.
- Advantages: Versatile and applicable in various contexts.
- Disadvantages: Difficult to construct bipolar adjectives.
What is a Semantic Differential Scale?
* A bipolar rating system that measures attitudes between two opposing concepts
* Typically uses a 7-point scale between contrasting adjective pairs
* Developed by Charles Osgood, George Suci, and Percy Tannenbaum in 1957
Key Characteristics
* Bipolarity: Uses opposing word pairs (e.g., Good-Bad, Fast-Slow)
* Versatility: Can measure multiple dimensions of attitudes
* Middle Point: Includes a neutral center point
Advantages
* Visual Appeal: Easy to understand and complete
* Rich Data: Captures nuanced attitudes across multiple dimensions
* Intuitive Design: Natural way to express contrasting feelings
Applications
* Brand Image Research: Comparing brand perceptions
* Product Development: Evaluating design features
Itemized Scale:
Stapel Scale
- Basic Characteristics: A unipolar ten-point scale ranging from -5 to +5, without a neutral point (zero).
- Example: Used for measuring attitudes and images.
- Advantages: Easy to construct and administer over the telephone.
- Disadvantages: Can be confusing and difficult to apply.
What is a Stapel Scale?
* A unipolar rating system used to measure attitudes and perceptions
* Features a single column of numbers, typically ranging from -5 to +5
Key Characteristics
* Simplicity: Single-column format for ease of response
* Flexibility: Adaptable to 5, 7, or 10-point scales
* Unipolar Nature: Measures intensity of a single attribute, positive or negative
Advantages
* Effective for Comparisons: Ideal for assessing attitudes towards multiple attributes
* Clear Interpretation: Straightforward data analysis, indicating positive or negative perceptions
* Reduces Neutral Responses: Lacks a neutral middle point, encouraging more decisive responses
Applications
* Primarily used in marketing and consumer research
* Suitable for measuring brand perception, product attributes, and customer satisfaction
Simple Random Sampling
Probability Sampling Methods
- Every member has an equal chance of selection
- Example: Drawing names from a hat to select lottery winners
Systematic Sampling
Probability Sampling Methods
- Selects members at regular intervals from an ordered list (much more efficient than simple random sampling)
- Example: Choosing every 10th customer entering a store for a survey
Stratified Sampling
Probability Sampling Methods
- Divides the population into subgroups and samples each subgroup
- Example: Polling different age groups to understand voting preferences
Cluster Sampling
Probability Sampling Methods
- Divides the population into clusters, randomly samples clusters, then samples all members within these clusters
- Example: Surveying all households in randomly selected city blocks
Convenience samples:
Nonprobability Sampling
Samples drawn readily available to the researcher
* Quick and cost-effective
* Example: Surveying local mall shoppers
Judgmental Samples:
Nonprobability Sampling
Samples selected based on researcher’s expertise
* Relies on subject matter knowledge
* Example: Expert panel for product feedback
Quota Samples:
Nonprobability Sampling
Samples reflecting certain characteristics in set proportions
* Controls for specific attributes
* Example: Percentage of females as representing the typical customer of an online shop
Snowball Samples:
Nonprobability Sampling
Samples accruing through participant referrals
* Useful for hard-to-reach populations
* Example: Customer satisfaction of an elite country club
