Table 1 Instruments measuring motivation to change in eating disorders

Assessment of the stages of change according to transtheoretical model of behaviour change

Readiness and Motivation Interview for eating disorders (RMI)

[19, 20]

Semi-structured interview

Four subscales:

N = 99 AN, BN, < EDNOS adults (inpatients)

Interrater agreement:

CV: Significant correlations with URICA and PCQ

‘Precontemplation’, ‘Contemplation’, ‘Action’, ‘Internality’

95.6% - 97.4%

DV: Non-significant correlations with age, socio-economic status, BMI, and social desirability^a

α = .63 - .84

PV: Prediction of anticipated difficulty of completing tasks related to eating disorder recovery^b, completion of recovery activities^c, decision to enrol in treatment, and dropout

[21]

Two subscales:

N = 65 AN, BN, < EDNOS adolescents (inpatients and outpatients)

Interrater agreement:

CV: Significant correlations with URICA and PCQ

‘Precontemplation’, ‘Action’ (as the internal consistency for ‘Contemplation’ and ‘Internality’ was unacceptably low)

90.3% - 97.9%

DV: Non-significant correlations with age, socio-economic status, BMI, and social desirability^d

α = .19 - .79

PV: Significant correlations with anticipated difficulty of completing tasks related to eating disorder recovery^b; prediction of completion of recovery activities^c

Anorexia Nervosa Stages of Change Questionnaire (ANSOCQ)

[22]

Self-report questionnaire (20 items)

Each item is regarded separately and a total score can be calculated.

N = 71 AN adults and adolescents (inpatients)

r _tt  = .89 α = .90

CV: Significant correlations with URICA; negative correlations of ANSOCQ total score with EDI-2

DV: Non-significant correlations of ANSOCQ total score with social desirability in adults^a, but positive correlations with social desirability in adolescents^e

PV: Prediction of weight gain during treatment by ANSOCQ total score; significant correlations between ANSOCQ total score at commencement of treatment and EDI-2 at discharge

[23]

N = 44 AN adults and adolescents (inpatients)

CV: Positive correlations of ANSOCQ total score with self-efficacy^f, DB subscale ‘Burden’, and negative correlations with DB subscales ‘Benefits’ and ‘Avoidance Coping’, and CSS total score

[24]

N = 70AN adolescents (inpatients and outpatients)

r _tt  = .90α = .94

CV: Negative correlations of ANSOCQ total score with EDI-2 and BDI

Bulimia Nervosa Stages of Change Questionnaire (BNSOCQ)

[25]

Self-report questionnaire (20 items)

Each item is regarded separately and a total score can be calculated.

N = 30BN adolescents (inpatients)

r _tt  = .94α = .93

CV: Negative correlations of BNSOCQ total score with BDI-2 and EDI-2

DV: Non-significant correlations of the BNSOCQ total score with BMI and illness duration, but positive correlations with age

Eating Disorders Stage of Change (EDSOC)

[26]

Self-report questionnaire (8 items)

Each symptom domain is regarded separately.

N = 145AN, BN, < EDNOS adults and adolescents (inpatients and outpatients)

r _tt  = .55 - 1.00α = .33 - .78

CV: Positive correlations of ‘Restrict’, ‘Diet Pill Use’, and ‘Fast’ with URICA

DV: Non-significant correlations with BMI, but negative correlations of ‘Fast’, ‘Restrict’, ‘Purge’, ‘Laxative Use’, and ‘Diet Pill Use’ with BSQ and positive correlations of ‘Purge’, ‘Laxative Use’ and ‘Diet Pill Use’ with age

Motivational Stages of Change for Adolescents Recovering from an Eating Disorder (MSCARED)

[27]

Questionnaire filled out together with an interviewer

Motivation for change and, if the youth is in action or maintenance phase, actions currently undertaken are rated.

N = 34AN, BN, < EDNOS adolescents (outpatients)

r _tt  = .92

CV: Positive correlations of the youth’s self-reported stage of change with the interviewer’s and the mother’s; lower EDI-2 and CDI scores in higher phases

(N = 16)

DV: Non-significant correlations with the diagnostic category, and with initial or final BMI levels

Readiness and Motivation Questionnaire (RMQ)

[28]

Self-report questionnaire (5 items for each of 12 symptom domains)

Two motivational stage scores (‘Precontem-plation’, ‘Action’) for each of four symptom domains; locus of control (‘Internality’, ‘Confidence’)

N = 207AN, BN, < EDNOS adults (outpatients)

r _tt  = .62 - .81α = .55 - .80

CV: Positive correlations of EDI with ‘Precontemplation’ and negative correlations with ‘Action’ and ‘Confidence’. Significant correlations with URICA and RMI.

DV: Non-significant correlations with BMI, self-efficacy^c, and social desirability^a; negative correlations of ‘Confidence’ with age

PV: Significant correlations with the anticipated difficulty of recovery activities^f and completion of recovery activities^d

Decisional Balance Scales

Decisional Balance Scale for Anorexia Nervosa (DB)

[29]

Self-report Likert scale (72 items)

Three subscales:

N = 246AN adults (inpatients and outpatients)

r _tt  = .64 - .71α = .88 for each of the three subscales

‘Burdens’, ‘Benefits’, ‘Functional Avoidance’

[30]

N = 80AN < EDNOS adults (outpatients)

CV: Positive correlations of ‘Burdens’ with PCQ; non-significant correlations of ‘Benefits’ and ‘Functional Avoidance’ with PCQ

DV: Non-significant correlations with social desirability^a, socio-economic status, and BMI; significant correlations of ‘Functional Avoidance’ with age

Pros and Cons of Anorexia Nervosa scale (P-CAN)

[31]

Self-report Likert scale (50 items)

Six pro-scales:

N = 233AN adults (inpatients and outpatients)

r _tt  = .60 - .85α = .52 - .78

CV: Positive correlations of P-CAN pro-scales ‘Appearance’, ‘Communicate Emotions/Distress’, ‘Fitness’, and ‘Safe/Structured’ with EDI; negative correlations of the P-CAN con-scale ‘Hatred’ with EDI

‘Safe/Structured’; ‘Appearance’; ‘Fertility/Sexuality’; ‘Fitness’; ‘Communicate Emotions/Distress’; ‘Special/Skill’

DV: Non-significant correlations of the P-CAN subscales with BMI

Four con-scales:

‘Trapped’; ‘Guilt’; ‘Hatred’; ‘Stifles Emotions’

[32]

N = 48AN adolescents (inpatients and outpatients)

α = .73 - .97

CV: Positive correlations of the P-CAN pro-scales ‘Communicate Emotions/Distress’, ‘Special’, ‘Safe/Structured’ with EDE-Q global score; positive correlations of P-CAN con-scales with EDE-Q total score

DV: Non-significant correlations of P-CAN subscales with BMI

Pros and Cons of Eating Disorders scale (P-CED)

[33]

Self-report Likert scale (72 items)

Subscales of P-CAN and four additional ones: pro-scales:

N = 202past or current diagnosis of AN or BN adults (outpatients)

None reported

DV: Significant differences between patients with AN and BN on P-CED subscales ‘Safe/Structured’ (AN > BN), ‘Special/Skills’ (AN > BN), ‘Fitness’ (AN > BN), ‘Fertility/Sexuality’ (AN > BN), ‘Eat but Stay Slim’ (AN < BN), ‘Guilt’ (AN > BN)

‘Boredom’; ‘Eat but Stay Slim’; con-scales: ‘Negative Self-Image’; ‘Weight and Shape’