Are there any problems with this conclusion? Read details as under.?

Question

Teen Challenge (A drug and alcohol residential treatment program) conducted an "Alumni Survey" to determine how successful their program was a treating drug addiction. They randomly selected 50 Alumni and mailed them a survey. 25 alumni responded with a completed survey. From this Teen Challenge concluded that their success rate at treating alcohol and drug addiction is 67%. Are there any problems with this conclusion?

Answer 1

The sample size is not nearly big enough... And there is a question about the reliability of voluntary surveys. (The type of people who answer voluntary surveys are usually very happy or very dissatisfied with the product, so your results could be skewed.)

Answer 2

Yes, there are problems. The other 25 who did not respond probably did that because they were not benefited. So, probably only 15 out of 50 got the benefit and that is only 30%.

In a sample survey, a large percentage of no-responses indicates that the sample may not represent the population adequately. Personal interviews will be the better option.

Answer 3

There's more than self-selection bias involved here.

Half of the answer is that you can't get 67% out of 25 people. Math forbids. It would mean that 16.75 people answered yes. (You can get 68% if 17 people answer yes.)

The other half of the answer is that unlicensed residential treatment centers for "troubled teens", such as Teen Challenge, often fabricate numbers, which is apparently what they're doing here. So odds are the data aren't any good to begin with.

Answer 4

subject is that non-responding 25 alumnis are lots greater possibly to stay addicts - too embarassed or stoned to respond, or moved without setting up mail forwarding. additionally, from a in basic terms mathematical attitude, per the place ultimate 25 are, we are able to basically say that fulfillment fee is between 33% and 80 3%. There are some papers and books via Charles Manski on that.