1.4 Threats to the quality of the scientific process

Kovacs and colleagues (2021) conducted a research project to elucidate the origins and causes of the most frequent mistakes committed by researchers that damaged the quality of the scientific process. Using a mix of quantitative and qualitative approaches to data collection, they surveyed 488 researchers publishing in psychology journals between 2010 and 2018. They found four supra-ordinal categories or metagroups and their corresponding causes (Table 1.2).

Table 1.2: Metagroups for Mistake Causes

Metagroup	Cause group
Poor project preparation or management	Bad or lack of planning
	Bad or lack of standards
	Bad skill management
	Miscommunication
	Failure to automate an error prone task
	Time management issue
External difficulties	High task complexity
	Technical issues
Lack of knowledge	Lack of knowledge/experience
Personal difficulties	Carelessness
	Inattention
	Lack of control
	Overconfidence
	Physical or cognitive constraints
Note. Table reproduced from Kovacs, Hoekstra, and Aczel (2021).

Figure 1.3 shows the mistakes reported by Kovacs et al. (2021) organized by the stages of research data management. To mitigate these mistakes different solutions have been proposed. For example, to define the data properly and to avoid ambiguous naming or to improve poor documentation practices, the use of codebooks, data management plans, and transparent research workflows is advisable (e.g., see Arslan, 2019). Similarly, to avoid wrong data processing and analysis, programming errors or loss of materials/data, using statistical code languages such as R or Python with embedded comments and storing the code in public repositories like Open Science Framework, Figshare, or GitHub is recommended (Klein et al., 2018).

Figure 1.3: Threats to the quality of the scientific process (Kovacs, Hoekstra, \(\&\) Aczel, 2021)