Both Qualitative and Quantitative
We interview individuals, not just "the sample."
Qualitative because our self-norming Linescale focuses on benchmarked valid scoring and underlying motivations for each individual. Quantitative because our expert system allows us to quickly interview many individuals and graphically display analyzed results immediately.
Methodology: Linescale uses psychometric techniques described in Nobel Prize winner Daniel Kahneman's book, Thinking, Fast and Slow. First, REFERENCE ANCHORS are established by respondent's identifying and rating their two favorite alternatives in the space.
Then, the idea is tested in two exposures. The first exposure is designed to measure the "System 1", or fast, intuitive faculty of our mind. The second exposure is designed to engage "System 2", or the slower, more thoughtful and logical faculty of our mind. Each item is rated on the same Linescale. Respondents see their earlier ratings.
NEW UNIQUE R&D TOOLS
- Linescale allows direct comparison with competitive alternatives, comparison with anchors
- Every study benchmarks all competitive alternatives
- Linescale allows direct comparison with expectation or initial impression or current impression
- Rating many items on the same Linescale gives more statistical power - paired comparisons
- Linescale rank order data (multiple paired comparisons) improves discrimination, significance
- Linescale allows fine discrimination inside the guides or "boxes," no matter the number of scale point guides
IMPROVEMENTS TO TRADITIONAL R&D TOOLS
- More granular scaling, sharper discrimination - an improved Likert scale
- Sharper, more reliable discrimination for Multivariate Analysis, Cluster or Factor Analysis
- Powerful individual respondent dependent variables
- Paired Comparison data in addition to six-point means and "box" score distribution
LINESCALE DOES BENCHMARKED TESTING NOT SURVEYS
- Direct measure versus expectation and competition allows scoring each respondent
- Preference Segmentation of each individual respondent shows size of target population
- Recommendation rating plus key variable ratings improves algorithm score
- Respondents see their scores, get to agree or disagree
- Respondents "explain" why they rated, Verbatim plus selection of key reasons for rating
- Preference Segmentation allows direct Driver Analysis to understand reasons for preference
- Subtle Driver Analysis clearly presented leads to key insights
R&D IN A NEW KEY YIELDS BETTER INFORMATION
- Your customers and respondents love the Linescale interview. Let us show you the data to prove it!
- Conversational style, replay and review of ratings leads to engaged respondents, better data
- Respondents rate the interview to insure accuracy, representativeness and engagement, and they rate it highly!
- Rating on Linescales much preferred by respondents to "Grids" - less cheating, fewer aborts and a higher voluntary completion rate
- Pleasant interview seen as a "feature" by customers, not a task
LINESCALE CONJOINT IS THE SIMPLE ALTERNATIVE
- Here's a summary of the differences between Linescale Conjoint and Traditional Conjoint: Linescale Conjoint uses direct paired comparisons on multiple scales. Traditional conjoint does many individual paired comparisons for preference of A vs. B, A vs. C, A vs.D, A vs.E, A vs. F, B vs. C, B vs.D, etc. Linescale Conjoint does direct paired comparisons of each with all. And, because of the efficiency of the Linescale we are able to do these direct comparisons on multiple scales rather than simple preference alone. This gives us the direct and real paired comparison of each item to each other item. Linescale also displays the average paired comparison rating of each item versus all other items. Plus, the scale rating and "box" scale values of each item are shown as byproducts of the measurement.
- In essence, Linescale Conjoint measures payoff preferences on absolute scales versus known benchmarks. The features are (1) Relative Linescore - percent of sample who are Acceptors, Borderline, Indifferent and Rejectors for each item, (2) paired preference for each item versus a known benchmark - usually the respondent's current favorite product, (3) average paired preference of each item versus all other tested items, (4) individual paired preference for each item versus all other individual items. And, (5) not only for overall preference, but for three additional key variables related to likely purchase and repeat purchase.
- Linescale assumes intelligence and insight on the part of the product developer who constructs realistic alternatives. We prefer to limit assumptions and directly pit the items against each other and against known controls. Instead of simulations based on statistical assumptions, we directly calculate the number of individuals in a target market who are likely purchasers of each alternative (Acceptors).
- When to use each?
Traditional Conjoint can be very helpful if you know little or nothing about a market.
Linescale Conjoint assumes informed product developers understand the category fairly well.
| || User Guide for Clients |