Mar 27, 2026
MIPPE Scale Validation Protocol
- Emmanuelle Dufait1,
- Sophie Buchheit2,
- Bernard Kabuth3,
- Fabienne Ligier4,
- Alexis Renard5
- 1University of Lorraine — InterPsy Laboratory (EA 4432);
- 2University of Lorraine, Laboratoire EA 4432 InterPsy (InterPsy Laboratory), Nancy, Pôle du Grand Nancy, Centre Psychothérapique de Nancy (Nancy Psychotherapeutic Center), Laxou, France;
- 3University of Lorraine, Laboratoire EA 4432 InterPsy (InterPsy Laboratory), Nancy, Pôle Universitaire de Psychiatrie de l'Enfant et l'Adolescent (PUPEA), Centre Psychothérapique de Nancy (Nancy Psychotherapeutic Center), Laxou, France;
- 4University of Lorraine, Laboratoire UMR 1319 INSPIIRE, Pôle Universitaire de Psychiatrie de l'Enfant et l'Adolescent (PUPEA), Centre Psychothérapique de Nancy (Nancy Psychotherapeutic Center), Laxou, France;
- 5University of Lorraine, University of Lorraine, Second Year of Master's in Applied Mathematics – Specialization: Probability and Applied Statistics, Technopole, Metz, France
- PERL
Protocol Citation: Emmanuelle Dufait, Sophie Buchheit, Bernard Kabuth, Fabienne Ligier, Alexis Renard 2026. MIPPE Scale Validation Protocol. protocols.io https://dx.doi.org/10.17504/protocols.io.8epv5y59dl1b/v1
Manuscript citation:
Version 1.0 — 2024
License: This is an open access protocol distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Protocol status: Working
We use this protocol and it's working
Created: March 27, 2026
Last Modified: March 27, 2026
Protocol Integer ID: 314029
Keywords: psychometric validation procedure of the mippe, child interactions in children, child interaction video, quality of early parent, child interaction, measure of early parent, early childhood clinician, mippe scale validation protocol, early parent, psychometric validation procedure, accessible to early childhood clinician, early childhood professional, early childhood professionals across multiple discipline, mippe scale validation protocol this protocol, observational rating scale, mesure des interactions précoces parent, item observational rating scale, transparent reporting of the validation methodology, mippe, parent, children, validation methodology, video corpus preparation through statistical analysis
Abstract
This protocol describes the psychometric validation procedure of the MIPPE (Mesure des Interactions Précoces Parent-Enfant / Measure of Early Parent-Child Interactions), a novel 9-item observational rating scale developed to assess the quality of early parent-child interactions in children aged 4 to 24 months. The MIPPE was designed to be accessible to early childhood clinicians in routine practice without requiring extensive training. The validation was conducted using a corpus of 228 parent-child interaction videos collected within the PERL randomized controlled trial (Petite Enfance, Recherche-action en Lorraine), rated by 143 early childhood professionals across multiple disciplines. This protocol covers the full validation procedure, from video corpus preparation through statistical analysis, and is intended to support reproducibility and transparent reporting of the validation methodology.
Materials
**Video Corpus**
- 228 standardized video recordings of parent-child dyadic interactions, collected during the PERL program's evaluation home visits
- Videos were recorded at two developmental time points: 4 months (N = 121 videos) and 24 months (N = 107 videos)
- Recording contexts included: free play, diaper change, bathing, and feeding
- Participants: 123 parent-child dyads (intervention group N = 62; control group N = 61)
- Data collection period: January 2019 – February 2024
- All families provided written informed consent for video recording under the PERL ethics protocol (CPP Nord-Ouest IV, IDRCB: 2017-A00896-47; CNIL Decision DR-2018-10)
**MIPPE Scale**
- 9-item observational rating scale, scored 0 to 3 per item (total score: 0–27)
- Items cover: quality of visual interactions (item 1), quality of vocal interactions (item 2), body position and physical contact (item 3), expression of affects (item 4), engagement and reciprocity (item 5), supportive parental presence/sensitivity (item 6), parental intrusiveness (item 7), evaluator's subjective state when observing the child (item 8), evaluator's subjective state when observing the parent (item 9)
- A detailed scoring manual accompanies the scale, specifying behavioral anchors for each rating level (0, 1, 2, 3) for each item
- Training time: approximately 2–3 hours using the manual and practice videos; no formal certification required
**Online Rating Platform**
- A dedicated, password-protected website was developed by a professional IT developer (May–October 2024)
- Access was restricted to invited participants via a unique URL, ensuring voluntary and informed participation
- Platform structure: (1) welcome page and scale presentation; (2) detailed instructions and scoring guide; (3) random video assignment; (4) rating interface with direct data transmission; (5) sociodemographic data collection (profession, seniority, clinical population); (6) satisfaction questionnaire; (7) closing page with study information
- The platform has since been taken offline to protect participant data
**Statistical Software**
- R version 4.3.2
- Packages: psychometric (reliability), lavaan with DWLS estimator (factor analyses), psych (internal consistency), stats (correlations, ANOVAs, t-tests)
- Bootstrap resampling was applied to estimate robust confidence intervals
Troubleshooting
Validation Procedure
Preliminary Rating of the Full Video Corpus: The lead researcher (E. Dufait) rated all 228 PERL videos using the MIPPE scale. This preliminary rating step served two purposes: (1) familiarization with item functioning and identification of potential ambiguities leading to final scale refinements; (2) classification of videos into three quality categories ('good', 'average', 'poor' interaction) to guide the selection of representative videos for the psychometric validation sample.
Selection of Representative Videos for Validation: Six videos were selected from the 228-video corpus by the research team (including 4 experienced early childhood clinicians) to represent the full range of interaction quality:
2 videos rated as 'good' interaction (1 at 4 months, 1 at 24 months)
2 videos rated as 'average' interaction (monitoring recommended)
2 videos rated as 'poor' interaction (intervention recommended)
Selection criteria included representativeness of interaction quality levels, video technical quality (image stability, audio clarity), and diversity of caregiver-child dyad characteristics.
Recruitment of Professional Raters: Early childhood professionals were recruited through professional networks, institutional contacts, and social media channels targeting the following socio-professional categories:
Physicians (psychiatrists, pediatricians, child psychiatrists, general practitioners, PMI physicians, midwives)
Nurses and child nurses (nurses, psychiatric nurses, pediatric nurses, advanced practice nurses, child nurses)
Paramedical professionals (speech therapists, educators, psychomotor therapists)
Psychologists and art therapists
Occupational therapists
Recruitment targeted a minimum of 100 raters to ensure adequate statistical power for factor analysis and reliability estimation, consistent with established guidelines for scale validation (5–10 participants per item minimum, with a minimum of 100 total).
Rating Procedure via Online Platform: Each participant was directed to the dedicated online platform via the unique URL shared through professional networks. Upon accessing the platform, participants:
Read the MIPPE scale presentation and scoring instructions
Were randomly assigned one panel of 6 videos (all at 4 months, representing good, average, and poor interaction quality)
Rated each video using the 9-item MIPPE scale
Provided sociodemographic information (profession, years of experience, type of clinical population)
Completed a brief satisfaction questionnaire
Each rater assessed the same panel of 6 videos. Participation was entirely voluntary. No financial compensation was provided. Anonymity was guaranteed throughout the data collection process.
Data Collection and Management: Rating data were transmitted directly from the online platform to the research team's secure email. Data were then entered into a structured database and pseudonymized. No nominative data were retained. Data from raters who did not complete all 6 ratings were excluded from analysis.
Final sample: N = 143 raters who provided complete ratings across all 6 videos, yielding 858 individual video ratings (143 × 6).
Statistical Analyses
Descriptive Analyses: Distribution of responses for each MIPPE item across the 143 raters, tests of normality (Shapiro-Wilk) at item level and total score level, homogeneity of variances (Levene's test).
Content Validity — Internal Consistency: Cronbach's alpha (target: α 3e 0.70), McDonald's omega (more appropriate for non-normal Likert data), corrected item-total correlations (threshold: r 3e 0.30).
Construct Validity — Factor Structure: Exploratory Factor Analysis (EFA) performed using a polychoric correlation matrix (appropriate for ordinal Likert data); Kaiser-Meyer-Olkin (KMO) index and Bartlett's test of sphericity to assess sampling adequacy; eigenvalue 3e 1 criterion for factor retention.
Confirmatory Factor Analysis (CFA): tested using lavaan package with DWLS estimator; fit indices evaluated: CFI (target ≥ 0.95), TLI (target ≥ 0.95), RMSEA (target ≤ 0.06), SRMR (target ≤ 0.08), chi-square p-value.
Discriminant Validity: Two-factor ANOVA (socio-professional category × interaction quality) to assess whether MIPPE scores differ across professional backgrounds. Expected result: significant effect of interaction quality (p 3c 0.05), non-significant effect of professional category (confirming cross-disciplinary usability). Levene's test for homogeneity of variances between professional groups. Seniority analysis: comparison of MIPPE scores across experience levels (threshold: 11 years).
Convergent and Divergent Validity: Convergent: Pearson correlations between MIPPE total scores and: maternal education level, 'desired child' variable, Revised Brunet-Lézine developmental scale (language subscale), ADBB (alarm détresse bébé) scale — assessed at 4 months (N = 121) and 24 months (N = 107).
Divergent: correlation between MIPPE scores and presence of addiction in the family (expected: negative correlation).
Clinical Threshold Determination: Comparison of multiple threshold combinations to optimize classification into three quality categories (poor / average / good). Optimization criteria: sensitivity (prioritized), specificity, and overall accuracy. Final thresholds: ≤ 15 (poor interaction, intervention recommended); 16–20 (average, monitoring recommended); 3e 20 (good quality).
Ethical Considerations
The original PERL research protocol received ethics approval from the Committee for the Protection of Persons North-West IV (CPP Nord-Ouest IV) on November 15, 2017 (reference IDRCB: 2017-A00896-47), with updated approval for substantial amendment N°7 on August 24, 2023. The PERL study also received authorization from the French National Commission for Information Technology and Civil Liberties (CNIL, Decision DR-2018-10, January 30, 2018).
All parent participants provided written informed consent prior to inclusion, including specific authorization for video recording and use for research, educational, and scientific presentation purposes. Professional raters provided implicit informed consent through voluntary access to the online platform. Rater data were fully anonymous.
Expected Outcomes
A validated 9-item observational scale for early parent-child interaction assessment (ages 4–24 months)
Unidimensional factorial structure with acceptable to excellent fit indices
High internal consistency (Cronbach's α ≥ 0.80)
Discriminant validity across 5 professional categories (no significant effect of profession on scores)
Convergent validity with established developmental and relational measures
Clinically usable thresholds for triaging interaction quality
Protocol references
Buchheit S, Kabuth B, Colombo M-C, Ligier F. Child Development and Early Interaction: PERL Research Protocol. Front Psychiatry. 2021;12:641468. doi: 10.3389/fpsyt.2021.641468