IDR Medical has successfully completed dozens of projects where conjoint analysis has been involved. We specify 4 aspects which require special attention in preparing conjoint studies for medical device & healthcare markets: sample, design, attributes and levels. Although we will discuss them one by one, they are all interconnected and influence each other.
Due to the specificity of the environment many studies with professionals (physicians or nurses and other clinicians) are limited by the number of respondents. A similar limit applies to the studies with patients, where we are looking for a very specific patient profile. A typical medical devices study sources 30 to 100 respondents per country, where in the context of FMCG markets, the number of respondents oscillates between 200 and 1000. Lower samples increase response error – and this needs to be compensated for. Paying extra attention to respondent profile and data quality is a must. All analysed respondents whose answers bear the marks of randomness, must be excluded from the study and replaced with valid specialists.
The two most common conjoint techniques ACBC and CBC have different requirements for minimum sample – ACBC works well with very low samples but requires larger commitment and time, while CBC is simpler and faster – but requires larger samples.
Generally, between thirty and sixty respondents per market/cell should be enough for investigational work and developing a hypothesis about a market.
In the medical devices market the number of tested products is lower than in the FMCG market (where the number of SKUs may be counted in dozens). In healthcare, the number of competitors is lower and usually 3-6 products are being tested (though they might be described by a larger number of attributes compared to typical FMCG goods). Both the sample and the number of attributes directly affect the conjoint design (the number of alternatives per choice, attributes differing per choice and the number of choices (tasks) per respondent). The whole design is a trade-off process between the statistical margin of error and the consistency of answers (variance of answers). More attributes provide more trade-off information, but also increase complexity and encourage the simplification of the strategy. More levels for each attribute provide more preference granularity, but larger sample sizes may be required to estimate the additional parameters. More alternatives per choice task also increase task complexity. More choice questions yield more data for a given sample size, but can contribute to respondents losing attention and an associated degradation of data quality. Because the overall precision depends on both response error and statistical error, larger sample sizes can help compensate for response error – but it’s often not possible due to limited respondent population and costs.
Product levels must be clear and unambiguous, so descriptions such as heavy/light or big/small should be avoided (though this regards conjoint in general not only in healthcare). It would be hard for someone to imagine whether 250g is a real problem comparing to 175g. Instead of that, the product should be described as a “cup of coffee weight” or “matchbox size”. The levels of pain are another example. Pain is usually difficult to describe, so doctors often use a smiley-face scale going from “very happy face – no pain at all” to “crying face – worst pain I’ve ever felt”. It still relies on individual judgements, but for analysis purposes, it works well enough.
Due to product complexity, descriptions of attributes should be issued before the conjoint starts. Unlike in FMCG markets, where most attributes are obvious (brand, package size), attributes of medical devices require clear descriptions to make sure that respondents fully understand them. Sometimes the description of an attribute is included in the study itself – as a reminder placed above tasks.
For more information and how we can help to solve your marketing problem please contact us at email@example.com or send us a message via our Contact Page.
If you wish to learn more about the nature of quantitative research, read our Introduction to Conjoint Analysis and An Overview of Conjoint Methodologies in Medical Device Markets. If your interests lie more on the qualitative side, then head to the Introduction to Qualitative Research in Medical Market, Projective Techniques in Concept Research or Depth Interview; The Answer to Complex Qualitative Issues piece. Or, if it’s the unmet need analysis that keeps you awake at night, read this article.
http://sawtoothsoftware.com, Technical Papers