The Q asked in the three polls: Did you have a serious problem with your [PS2, Xbox, GC] hardware?
As of now, the results are as follows:
PS2
Serious problems: 24 (52%)
No serious problems: 22 (48%)
Total # of cases: 46
Xbox
Serious problems: 7 (29%)
No serious problems: 17 (71%)
Total # of cases: 24
GameCube:
Serious problems: 2 (5%)
No serious problems: 37 (95%)
Total # of cases: 39
Total # of cases (all three polls): 109
Statistical limitations, strenghts and results:
1) Participants weren't randomly chosen, which results in an increased likelihood of bias for all socio-demographic data. The same bias is reduced for the target population, however, it is still very likely.
2) Because of 1) the poll isn't representative of the general population and/or the target population of game system owners, therefore the poll isn't generalizable.
3) The number of cases is low. However, a difference of means test which is appropriate for case numbers from around 25 to 100 showed across the groups and for the pair of groups within the target population statistical significant difference. The differences found exclude population difference based on sampling error. However, the commonly agreed upon statistical significance level of .5 cannot be tested due to the low number of cases.
4) The poll gives certainly a trend of hardware reliability. The found differences between the three systems are 1) substantial, and 2) not coincidental. The PS2 is certainly much more prone to perceived hardware problems than the Xbox and the GameCube. The qualitative posts for the PS2 which explained at least some of the problems, indicate clearly more serious hardware problems compared to the two other systems. The GameCube proved to be by far the most reliable piece of hardware.
6) Although the polling question wasn't specified (meaning of "serious", hardware") and didn't distinguish between different kinds of problems, this doesn't distort the trend across the systems at all. Potential misunderstandings, exaggerations, different assumptions, bad feelings for a specific system etc. are present for all three polls. It is unreasonable to assume that owners of a specific system are more intelligent, hate another system more, and/or are more or less prone to the above mentioned perception problems.
7) There is a bias in favor of the PS2. The PS2 was one year earlier released than the GC and the Xbox. Hardware problems are corrected over time for the subsequent versions of a system. If we assume that early childhood flaws of a system are corrected, the more recently released systems are disadvantaged. The fact that the oldest system, the PS2, turned out to be the most problematic piece of hardware by far, underlines the overall trend.
8) The target population is part of a videogame website dedicated primarily to video game collectors. It is reasonable to assume that videogame collectors are over-represented in the sample compared to the general population. Videogame collectors in general have experience and an incentive to treat game hardware carefully and appropriately. The findings of a PS2 prone to serious hardware problems is therefore even more troublesome. (unless we assume that the more careless non-collectors are clearly over-represented in the "serious problems" category of each system.)
9) Two of the systems (PS2, Xbox) have DVD and CD play capability, while the GC is only able to play game discs. The troublesome findings for the PS2 still hold up. 1) Compared to the Xbox, the PS2 still proved to be more unreliable, 2) If a combo ‘game player/DVD player' is offered as an incentive to sell a system and in all likelihood results in more sales, problems with the DVD player as part of the system can certainly count as a serious hardware problem.
Well, draw your own conclusions. The polls together with accumulating reports from friends, game mags, and store employees make the problem-ridden PS2 hardware certainly much more than just a rumor or a distorted picture of anecdotes, in my opinion. It's one significant step above anecdotal evidence for sure.
Hey, thanks for participating! 109 is much more than I expected, and statistically you can certainly work with such a number of cases. Pretty bad results for the PS2, eh?