The paper discusses various methods of determining mean temperature in climate studies and the consequences of applying particular methods. The conclusion is that the use of various mathematical formulas that describe mean temperature usually yields different results. In some cases, the differences can reach ten to twenty percent. Consequently, the results of different analyses are incomparable, even in engineering, and especially in research and scientific studies. Hence the postulate to avoid conducting comparative analyses with the use of data from different authors, if their temperature definitions are different, or if there is no precise information on the formulas used in the calculation.