Statistics humor is good humor.

Seeing that I have some sort of weird fetish about doing double blinded experiments, I did a test on 4 brands of orange juice from the store across the road.

At first I was planning to do this experiment like I usually do them, but during the start of testing something weird happened.

In the first test one of the juices got a score of 6 (1-10) but in the subsequent tests got 2.5 and 3. The reason for this is “The orange juice deviation”, a law of nature that states that “If you drink two different types of orange juice in a short time span, the second one will taste fucking weird”.

I tried rinsing with water between each test, but that did not help, so I had to do the testing another way to avoid interference between the different samples.

So to avoid the taste testing being influenced by the previous one, I drank 24 cups of orange juice over the course of 3 days with a minimum of 30 minute delay between tests. Thanks to my GF Stacy for helping with the blinding (she suspected it was all a ploy to make her get me juice).

TEST 2 DATA SET
1	lerum		6
2	sunniva		6.5
3	first price	7
4	sunniva		8
5	first price	6.5
6	first price	6.5
7	first price	7.5
8	eldorado	5.5
9	sunniva		7.5
10	first price	7.5
11	lerum		7
12	first price	7
13	lerum		6.5
14	sunniva		7
15	lerum		8
16	eldorado	6.5
17	lerum		8
18	sunniva		8
19	eldorado	6
20	lerum		6
21	lerum		7
22	eldorado	7
23	sunniva		8
24	sunniva		7.5

Test 1 – drinking 4 glasses of randomized juice.

Test 2 – drinking 1 glass of random juice.

So what’s the conclusion?

I’ve always been buying the Eldorado juice, however I clearly think Sunniva is the best (highest score in both tests). however the 1 Liter pack of Sunniva is almost twice the cost of the 2 Liter pack of First Price which is a very cheap brand. Despite that First Price did came in second and who doesn’t love a juice with a small standard deviation?