Internal consistency reliability of mental health questionnaires in college student athletes

To examine the internal consistency reliability and measurement invariance of a questionnaire battery designed to identify college student athletes at risk for mental health symptoms and disorders.

College student athletes (N=993) completed questionnaires assessing 13 mental health domains: strain, anxiety, depression, suicide and self-harm ideation, sleep, alcohol use, drug use, eating disorders, attention deficit hyperactivity disorder (ADHD), bipolar disorder, post-traumatic stress disorder (PTSD), gambling and psychosis. Internal consistency reliability of each measure was assessed and compared between sexes as well as to previous results in elite athletes. Discriminative ability analyses were used to examine how well the cut-off score on the strain measure (Athlete Psychological Strain Questionnaire) predicted cut-offs on other screening questionnaires.

Strain, anxiety, depression, suicide and self-harm ideation, ADHD, PTSD and bipolar questionnaires all had acceptable or better internal consistency reliability. Sleep, gambling and psychosis questionnaires had questionable internal consistency reliability, although approaching acceptable for certain sex by measure values. The athlete disordered eating measure (Brief Eating Disorder in Athletes Questionnaire) had poor internal consistency reliability in males and questionable internal consistency reliability in females.

The recommended mental health questionnaires were generally reliable for use with college student athletes. To truly determine the validity of the cut-off scores on these self-report questionnaires, future studies need to compare the questionnaires to a structured clinical interview to determine the discriminative abilities.