The study results are not very accurate. They made it with only 1000 US random users, and ~60% chose Bing more often, ~30% chose Google more often (on 10 queries).
See the following fictive example:
1000 people, 600 have the following score: 4 Bing, 3 Google, 3 Draw; 400 have the following score: 10 Google.
So, in my example, 60% of the people chose Bing more often and 40% chose Google more often. But, if you look at queries numbers, Google was chosen for 3600+10400=5800 queries, and Bing for 4*600=2400 queries, that's a 2:1 ratio, but in favor of Google.
See the following fictive example:
1000 people, 600 have the following score: 4 Bing, 3 Google, 3 Draw; 400 have the following score: 10 Google. So, in my example, 60% of the people chose Bing more often and 40% chose Google more often. But, if you look at queries numbers, Google was chosen for 3600+10400=5800 queries, and Bing for 4*600=2400 queries, that's a 2:1 ratio, but in favor of Google.