Education Endowment Foundation: Assessment of Children’s Emotional Skills

About the measure

Subscales

Emotion attribution accuracy; Emotion attribution biases: 1 for each of sad, happy, fear, anger, no feeling

Example

Kelly just finished coloring a picture. You tell her that it looks “nice.” “Do you think Kelly feels happy, sad, mad, scared, or

Link

https://www.cambridge.org/core/services/aop-cambridge-core/content/view/953D415A967CC9EEBF054F31EEE7C3EB/S0954579404044566a.pdf/childrens_emotion_processing_relations_to_emotionality_and_aggression.pdf

Key Stage

Key Stage 1, Key Stage 2

Domains

Emotional Intelligence (mixed and trait), Social and Emotional Competence

Rating Psychometric

Rating Implementation

Implementation

No. of itemsⓘNumber of items included in the measure

FormatⓘType of response scale

Nominal

RespondentⓘWho is asked to complete the measure.

Self

ScoringⓘStandardised: Can easily be summed or averaged Complex: Requires judgement or assessment on behalf of the scorer

Standardised

AgeⓘAge (in years) covered by the survey. Using the measure outside this age range may result in inaccurate information.

5 – 8 year(s)

Cost single purchaseⓘCost (if applicable) of purchasing the measure.

Free

Cost per childⓘDetails if there are additional costs per administration. This can vary, so if a cost is indicated, further contact with the publisher/author is recommended.

Free

Psychometric

UK normsⓘWhether the measure has been calibrated with a UK sample (as opposed to generating psychometric details from a Foreign (e.g. USA) sample).

Chronbach's alphaⓘAn estimate of the reliability of the measure. A higher value indicates a more reliable measure (0.7 is normally considered a good benchmark for reliability).

.46-.70

Text retestⓘA measure of reliability. This indicates how ‘stable’ responses to a measure are over time.

Not reported

Inter-rater reliabilityⓘWhere observations (e.g. teacher or peer report rather than self report) are used, this measure of reliability indicates how closely related each score is. A higher value suggests a more reliable measure.

Not reported

EFAⓘExploratory Factor Analysis – This is measure of the validity, showing how well the underlying structure of the domain is represented by the individual items. Higher values (factor loadings) indicate a good fit between items and the sub-domains of the measure.

Not reported

CFAⓘConfirmatory Factor Analysis – Similar to EFA, this measure of validity assesses the extent to which items fit pre-specified sub-domains of a measure. Higher values indicate a better fit.

Not reported

Criterion validityⓘAn assessment of how well the measures relates to other, well-validated assessments.

Not reported

Construct validityⓘFurther to the use of CFA and EFA, this reports any details demonstrating how well the measure is seen to represent the conceptual domain it comes from.

Emotion Attribution Accuracy correlated to Happiness measured as an aggregated index from a sociometric interview/Differential Emotions Scale, partial r= .19; The anger attribution bias was also related to an index of anger, partial r= .17 and peer-nominated sadness, partial r= .15; The fear attribution bias was related to teacher-reported fear, partial r= .16; The no feeling attribution bias was related to peer-nominated fear, partial r= .15

Predictive validityⓘThe extent to which the measure can be seen to predict other outcomes.

Not reported

ResponsivenessⓘAn indicate of how well a measure if able to detect change during a pre-specified time (if applicable).

Not reported

Floor/CeilingⓘIndicates whether there are any issues with ceiling (maximum scores) or floor (minimum scores) effects with large portions of particular groups of respondents.

Not reported

References

Schultz, D., Izard, C. E., & Bear, G. (2004). Children’s emotion processing: Relations to emotionality and aggression. Development and Psychopathology, 16(2), 371 – 387. doi:10.1017/s0954579404044566

Education Endowment Foundation: Assessment of Children’s Emotional Skills

About the measure

Subscales

Example

Link

Key Stage

Domains

Rating Psychometric

Rating Implementation

Implementation

No. of itemsⓘNumber of items included in the measure

FormatⓘType of response scale

RespondentⓘWho is asked to complete the measure.

ScoringⓘStandardised: Can easily be summed or averaged Complex: Requires judgement or assessment on behalf of the scorer

AgeⓘAge (in years) covered by the survey. Using the measure outside this age range may result in inaccurate information.

Cost single purchaseⓘCost (if applicable) of purchasing the measure.

Cost per childⓘDetails if there are additional costs per administration. This can vary, so if a cost is indicated, further contact with the publisher/​author is recommended.

Psychometric

UK normsⓘWhether the measure has been calibrated with a UK sample (as opposed to generating psychometric details from a Foreign (e.g. USA) sample).

Chronbach's alphaⓘAn estimate of the reliability of the measure. A higher value indicates a more reliable measure (0.7 is normally considered a good benchmark for reliability).

Text retestⓘA measure of reliability. This indicates how ​‘stable’ responses to a measure are over time.

Inter-rater reliabilityⓘWhere observations (e.g. teacher or peer report rather than self report) are used, this measure of reliability indicates how closely related each score is. A higher value suggests a more reliable measure.

EFAⓘExploratory Factor Analysis – This is measure of the validity, showing how well the underlying structure of the domain is represented by the individual items. Higher values (factor loadings) indicate a good fit between items and the sub-domains of the measure.

CFAⓘConfirmatory Factor Analysis – Similar to EFA, this measure of validity assesses the extent to which items fit pre-specified sub-domains of a measure. Higher values indicate a better fit.

Criterion validityⓘAn assessment of how well the measures relates to other, well-validated assessments.

Construct validityⓘFurther to the use of CFA and EFA, this reports any details demonstrating how well the measure is seen to represent the conceptual domain it comes from.

Predictive validityⓘThe extent to which the measure can be seen to predict other outcomes.

ResponsivenessⓘAn indicate of how well a measure if able to detect change during a pre-specified time (if applicable).

Floor/CeilingⓘIndicates whether there are any issues with ceiling (maximum scores) or floor (minimum scores) effects with large portions of particular groups of respondents.

References

Cost per childⓘDetails if there are additional costs per administration. This can vary, so if a cost is indicated, further contact with the publisher/author is recommended.

Text retestⓘA measure of reliability. This indicates how ‘stable’ responses to a measure are over time.