Tag Archives: playscan research

A shorter Self Test: Does not increase completion rate

Summary: Shortening the 16 statement Self Test within Playscan yields negligible improvement in completion rates. The length of the test is not a problem; players either drop off during the first couple of questions or complete the test.

 

A recurring concern about Playscan has been that 16 statements to consider in the Self Test may be too many. The player may grow impatient and abort the test, especially since the questions themselves can be sensitive and draining. We investigated whether a shorter introductory test with “gate questions” would increase the completion rate of tests.

 

When a player clicks into the Self Test, an introductory text is displayed. Here, the player is encouraged to consider all gambling, at all gambling sites, during the past three months. The player is then asked to consider 16 statements, one at a time.

 

playscan_selftest1

 

To investigate the usefulness of gate questions, the results from the Self Tests at Svenska Spel between 2014-07-04 and 2014-10-14 were analyzed. Statistics of these are presented below, showing the completion and drop-off rates.

playscan_selftest_drop_off_rate

 

Looking at the numbers, the completion rate is quite satisfactory; in particular 80% web completion. This high number is likely due to the curiosity that brought the player to Playscan in the first place, and the promise of self-assessment at the end of the process. Self tests in general tend to have a higher completion rate than surveys thanks to the intrinsic motivation behind doing them.

 

The majority of the players who drop off do so at the first question. We also see a difference between channels with a 10% drop-off rate on web and 23% on mobile. The higher drop-off on mobile is hardly surprising, given the users’ attention span in the mobile context.

 

Only 10% of the started tests are dropped between question 2 and 16, regardless of channel. Interesting to note is that the drop-off rate declines as the test continues.

 

This leaves us with a clear answer to the question of gate questions. We would have yielded only 4% more completed tests if the test consisted of four statements. This number is hardly worth chasing at the cost of the players spending less time contemplating their gambling habits or missing out on the nuances that the full 16 statements bring.

 

———

 

 

The research done at Playscan is not academically focused, but aimed at practical application.
We are pragmatists, knee deep in data to explore. Our mission is to help prevent problem gambling rather than to study it, so we spend our time chasing preventive effect wherever we sense it.
We value agility and adaptation.
Where the territory is uncharted, our guiding light is curiosity and making a difference. Our data is local. We sometimes see wildly varying player behavior between operators, not necessarily because the players are different, but because contexts and presentations are. We believe that the research community has lots to learn about the importance of things like wording and design, and what we say will often be framed to show this. Our findings reflect the everyday player experience. This is neither universal nor static. It can change and, more importantly, can be changed.
At the same time, we have the deepest respect for formal research and academics. We welcome critique of our findings, and hope that others find inspiration and ideas to bring into the academic world.  We are happy to help, and love to exchange experience and ideas. Give us a call if you would like to help out!