A new Slop University paper reports a pre-registered replication of the University's own Pick-Up Reliance Scale, finding the instrument as reliable as before and the association behind it a fraction of the size first reported, four months after a childcare provider began flagging families on the score.
A replication across 26 primary schools and 1,438 households reproduces the instrument's reliability but not the association it was adopted for
A twelve-item questionnaire went into an out-of-school-hours-care enrolment form in March. The University has now published what happened when somebody administered it again.
It Held in the First Sample, released today through the Office of Research Outputs, is a pre-registered replication of the Pick-Up Reliance Scale, an instrument developed at Slop University to score how far a household leans on a class group chat to run the school-gate handover. Across 26 primary schools and 1,438 households the scale’s reliability returned almost exactly as first reported. The association that carried it into service did not. The correlation the original put at 0.36 is 0.06 here, and the flag the provider raises at a score of 34 does not separate households on the handover failures it is meant to anticipate.
The reliability was never in question and it still isn’t. What we could not tell anyone until this week was whether the number carried the meaning the enrolment form has been giving it since March.
— Dr Ingrid Vasseur, Lecturer and Convenor, Futures in Committee
The team administered the twelve items unchanged across four school systems, paired them with the original’s four-week household diary and a second wave eleven weeks later, and set the resulting scores against the provider’s de-identified enrolment and incident records across eight terms.
Dr Thandiwe Solberg, Senior Lecturer and Convenor of Demo Quarter, put the interval at the centre of the exercise: “Somebody had to decide the instrument was ready. Nobody had to decide it was still ready.”
The University counts work of this kind as part of what it owes anything published under its own name.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.vjkmc1.