Beyond reproducibility, making a psychological application safe to use

The replication crisis inevitably deepens issues of trust in the methods and findings that the scientific community uses to advance our collective understanding of human behavior. In turn, any doubts over psychological research undermines the use of any products that it develops (e.g. assessment tools, scales, frameworks) outside of academia. With the help of the Implicit association test (IAT) this chapter shows how measurement criteria (reliability, validity) are used to determine metrics for determining a sound evidence base for the test. However, escaping the replication crisis on the grounds that a body of work is highly replicable and has high predictive validity is not sufficient when it comes to the issue of applicability. The chapter argues that without a good theoretical basis to secure construct validity the same research would be ruled out of being safe for use outside of the sciences. The reason for being even more stringent about the concepts being investigated is that the stakes are higher when the decisions based on them carry consequence in practical settings.