GEM Manifesto

10 principles we built our service on

In building MindCanvas, we created a new type of online research method. Since all things web-related have an acronym, we came up with one as well : GEM for Game-like Elicitation Methods. These principles below guided us while we built the software and conceptualized the service.

1. The overuse of Likert scales must stop. We understand that Likert scales have their uses, but every study to understand people does not need Likert scales. Likert scales are problematic in many ways – response biases, sheer rating boredom, statistical issues (ordinal or interval- depends on who you ask) are just some of them!
2. Research methods must allow you to explore the subjective, the implicit and the unconscious. Much of human thought lies below awareness. It’s high time we tapped into what’s beneath the surface.
3. Your research should not put participants to sleep! Products must not be built on the basis of data from online research where bored users are made to go through a screenful of multiple choice questions like mechanical robots.
4. Research should allow participants to express themselves using their own words and concepts. Research should elicit people’s concepts, categories and associations – it should gives us an understanding of the world from their perspective.
5. Online data collection must move beyond vanilla HTML. Rich interfaces should revolutionize online research just as they have revolutionized online maps (Google Maps), and Email (Gmail, Yahoo Mail).
6. Products that are used across cultures should not be researched and tested only in Silicon Valley, or in the home base of the company. International products need international user research.
7. Good research should not be the prerogative of people with PhDs. What’s important for the researcher is to be good at garnering design insights. Many of the challenges in good research design and statistical analysis, e.g., “How many users do I need?” “Is this study design appropriate?” “What statistical analysis do I need?” can be outsourced to an intelligent system or to an outside expert.
8. Deliverables should inspire designers. Deliverables should reveal insights garnered through statistics and data mining, in a visual manner. Researchers / designers should be able to play with the research results using intuitive, self-explanatory visualizations that capture the full richness of the data.
9. Deliverables should persuade business stakeholders. The research method should not become the point of contention. The focus should be the research findings and the product design, not the method itself. Large sample sizes and statistical analysis might help in persuading business stakeholders.
10. Participating in research should have the easy playfulness that we expect from games. Think of how much fun it is to play scrabble. Now, imagine scrabble redesigned as an HTML interface. Imagine, if research methods could be fun and engaging like scrabble.

Go back to top, learn about our research methods or check out the deliverables.

Have questions? Call 1-650-564-0000 or email support@themindcanvas.com for answers, to find out about purchasing MindCanvas or to get access to live demo!