Presenting dynamics of social phenomena: should we use absolute, relative or time differences?

Empirical studies within social sciences face an important decision about how to express key findings to the target audience. Simplicity is an important selection criterion here, because the findings need to be conveyed in an efficient manner (i.e., briefly and concisely), but also because stakeholders (e.g., policy makers, the media, general public) are heterogeneous in their methodological backgrounds. The corresponding ways of measuring thus need to be not only exhaustive and message-delivering but also simple and intuitively understandable. This is particularly important when dynamics in time are discussed. There, most typically, either absolute or relative differences are used. This review paper critically elaborates these two popular measures and, in addition, discusses the alternatives of time distance and time step. The paper demonstrates that even in simple linear examples, the results of these four types of measures may sharply contradict. The empirical example of the digital divide is also elaborated, which illustrates many tempting possibilities for biased, one-sided interpretations that match the needs of certain stakeholders. Finally, the paper alerts users about possible misleading conclusions and suggests comprehensive treatments, using several measures simultaneously. Copyright Springer Science+Business Media Dordrecht 2014