In 1975, the average daily maximum temperature at San Francisco airport was 63.7 degrees, with an SD of 8 degrees. This average temperature value was a little on the low side when compared with recent history: over the previous decade the corresponding average was 65.1 degrees. Somebody wonders if this difference is significant.
The standard machinery of significance testing either works or fails in a given situation, and when it fails it can do so for one of two reasons: conceptual (the whole idea of significance testing in this case doesn't make sense), or technical (the idea is sensible enough, but the technical assumptions underlying the usual tests do not hold).