Contemporary studies of long-term changes in the ionosphere stem mostly from the suggestions by Roble and Dickinson  that “global warming” in the lower atmosphere is accompanied by “global cooling” of the thermosphere, and subsequently, from the suggestion by Rishbeth  that the resulting thermal contraction lowers the height of the ionospheric F2-peak hmF2. The subject is attractive to study because decades of ionosonde data exist from dozens of stations worldwide, the data are well organized in a consistent format, and the analysis requires no great computing power. However, the trends from different stations are far from consistent and often show interruptions or reversals. There are tantalizing details, such as opposite trends of rising hmF2 at places east of longitude 30°E, falling hmF2 west of 30°E [Bremer, 1998]. We have to ask: Are these real? Indeed, we should ask the more general question: “What is needed to make reported trends “believable”?