The effect of prior record in sentencing research: An examination of the assumption that any measure is adequate