Meta-science in public administration : three meta-strategies for testing relationships in public service motivation theory

Abstract Statistically over 50% of assumed true findings in published quantitative research is most likely false (Ioannidis 2005). The field of public administration is no different in this respect. Within this behavioral PA... [ view full abstract ]

Abstract

Statistically over 50% of assumed true findings in published quantitative research is most likely false (Ioannidis 2005). The field of public administration is no different in this respect. Within this behavioral PA field, the topic of public service motivation (PSM) seems to be particularly prone to have an increased likelihood of biased results in terms of ‘true knowledge’ when looking at the field criteria as stated by Ioannidis (2005). It is a hot topic, not just in terms of publications (Ritz et al. 2016), but also in terms of citations (Vandenabeele and Skelcher 2015), with multiple teams working on similar research questions. Next to this, there is substantial flexibility regarding conceptual or operational definitions (Perry and Vandenabeele 2015; Vandenabeele et al. 2017). Furthermore, there seems to be some prejudice towards its validity, with scholars sometimes even divided in believers and non-believers (Vandenabeele et al. 2017). Finally, PSM research often is based upon low-power studies (however, more recent work is often based upon larger samples). It therefore would be particularly suited to expose to a set of meta-strategies, to cast more light on the true nature of the relationships in PSM theory.

This research develops three meta-strategies as a solution to deal with these false positives and applies this to the topic of public service motivation (PSM), which is the motivation people have to contribute to society – a core variable in the field of public administration. Two common assumptions about PSM are 1. that activation increases PSM levels and 2. that fit between the environment and PSM increase performance. These will be reviewed by means of these meta-strategies. The first meta-strategy is a classic meta-analysis based upon published and non-published studies. The second research strategy involves a replication (‘Many Labs’) of the same two experiments. In every country involved in the study (UK, D, NL), each experiment will be conducted twice, in order to control for country effects. The third meta-strategy is crowd-sourcing, for which a data set will be collected (three waves, multiple measures and correlates). This data set will then be subsequently analyzed by 40 different research teams in order to estimate a ‘crowd-sourced effect’. The contribution of the proposal will be as well on the more robust meta-findings on PSM, as well as on the variables influencing the meta-results of various strategies in order to get a better grasp on the properties of these strategies.