Stanford-Binet Intelligence Scales (SB‑5)

EYFS Early Learning Goals

The specific Early Learning Goal(s) from the 17 in the EYFS that the assessment most closely aligns with.

Early Years Evidence Store Approaches

The EEF Early Years Evidence Store approach(es) with which the outcome measure is most closely aligned.

Availability in national administrative datasets

Whether data from this assessment are collected and available within national administrative systems, such as the Department for Education’s National Pupil Database.

No

Administration format

Whether the measure is administered to an individual or a group.

Individual

Respondent

Whether the measure collects responses from the child, parent/​carer, or teaching professional.

Child

Response mode

The method through which responses are elicited or recorded during assessment administration, for example, behavioural observation, oral responses, manual tasks etc.

behavioural interaction

Item response type

The format used for recording each item’s response, for example, multiple choice, open-ended, rating scale, task completion, timed response etc.

Performance/​Task success

Prior qualifications/training required

Prior qualifications/​training required to administer the measure.

An adcanced degree in relevant field or a bachelors degree with additional training. Standards of practice for which professionals may engage in assessment vary greatly within different countries, states, and localities.

Ease of scoring

The nature of scoring the measure, including whether scoring requires expert judgement or uses complex calculation methods, where such information is available.

Easy. The SB‑5 can be scored by hand using the record form and examiner’s manual. The SB‑5 generates standard scores for each subtest, plus Composites for Verbal IQ; Nonverbal IQ; Brief IQ; Full Scale IQ; and five Factors – Fluid, Knowledge, Quantitative, Visual-Spatial, and Working Memory; also, optional criterion-referenced Change-Sensitive scores.

Availability of guidance

Availability of guidance for administering and scoring the measure.

Interpretive Manual, Examiner Manual, and Technical Manual available.

Time required

An estimate of the total time (in minutes) needed to administer the measure. Where reported, this includes both preparation and scoring time in addition to direct administration.

45 to 75 minutes.

Cost

Kits start from $505.00 with record forms at $93.00 for a pack of 25.

Availability

Available for purchase from the publishers’ website.

Standardised?

Whether a measure has been standardised on any population. A standardised measure is a measure that has been tested on a defined sample and has established norms, allowing scores to be compared against a reference population.

Yes

UK Standardisation sample size

UK standardisation sample size is the number of participants included in the study used to establish the test’s norms or reference scores.

Rating
N/A: Not standardised
N/A: Standardised with a non-UK population
Rating: 1 out of 5: <200
Rating: 2 out of 5: 200 – 999
Rating: 3 out of 5: > 1000

Construct validity

Construct validity shows how well a test measures the psychological construct it claims to assess.

Rating –
Rating: 1 out of 5 Limited: No evidence that items and test were developed based on theory and construct definition OR statistical tests (e.g. Factor Analysis, Item-test correlations) demonstrate lack of construct validity (see EFPA guidance for further detail). 
Rating: 2 out of 5 Moderate: Evidence that items and test were developed based on theory and construct definition.
Rating: 3 out of 5 Strong: Evidence that items and test were developed based on theory and construct definition AND statistical tests demonstrate construct validity and confirm construct validity (see EFPA guidance for further detail).

Criterion validity

Criterion validity assesses how well test scores correlate with relevant external criteria.

Rating –
N/A: Not criterion referenced.
Rating: 1 out of 5 Limited: (r < 0.20) correlation with the criterion measure OR small (< 100) sample size.
Rating: 2 out of 5 Moderate: (0.20 ≤ r < 0.50) correlation with the criterion measure AND adequate sample size (≥ 100).
Rating: 3 out of 5 Strong: (r ≥ 0.50) correlation with the criterion measure AND adequate sample size (≥ 100); OR Good (0.35 ≤ r < 0.50) correlation with the criterion measure shown in one large sample (> 200) or more than one adequate sized study.

Internal consistency

Internal consistency is a measure of how well the items work together to measure the same underlying construct.

Rating –
Not reported.
Rating: 1 out of 5 Limited: (r < 0.70) reliability coefficient OR small (< 100) sample size.
Rating: 2 out of 5 Moderate: (0.70 ≤ r < 0.80) reliability coefficient AND adequate sample size (≥ 100).
Rating: 3 out of 5 Strong: (r ≥ 0.90) reliability coefficient AND adequate sample size (≥ 100); OR Good (0.80 ≤ r < 0.90) reliability coefficient in one large sample (> 200) or more than one adequate sized study.

Test-retest reliability

Test-retest reliability shows how well a measure produces consistent scores when administered to the same individuals at different points in time.

Rating –
Not reported.
Rating: 1 out of 5 Limited: (r < 0.60) reliability coefficient OR small (< 100) sample size.
Rating: 2 out of 5 Moderate: (0.60 ≤ r < 0.70) reliability coefficient AND adequate sample size (≥ 100).
Rating: 3 out of 5 Strong: (r ≥ 0.80) reliability coefficient AND adequate sample size (≥ 100); OR Good (0.70 ≤ r < 0.80) reliability coefficient in one large sample (> 200) or more than one adequate sized study.

Inter-rater reliability

Inter-rater reliability measures how well different raters or observers produce consistent scores when assessing the same individuals or behaviours.

Rating –
Not reported.
Rating: 1 out of 5 Limited: (r < 0.60) reliability coefficient (pairwise correlation) OR small (<100) sample size.
Rating: 2 out of 5 Moderate: (0.60 ≤ r < 0.70) reliability coefficient AND adequate sample size (≥ 100).
Rating: 3 out of 5 Strong: (r ≥ 0.80) reliability coefficient AND adequate sample size (≥ 100); OR Good (0.70 ≤ r < 0.80) reliability coefficient in one large sample (>200) or more than one adequate sized study.

Non-UK standardisation sample

Standardisation sample size is the number of participants included in the study used to establish the test’s norms or reference scores.

SB‑5 is based on a representative sample of 4,800 individuals, including 1400 children aged 2 – 5. The stratification variables included sex, age, race/​ethnicity, socioeconomic level, and geographic region.

Validation sample

Standardisation sample characteristics are the key demographic and contextual details of the group used to develop the test norms (e.g. age range, gender balance, socioeconomic background, geographic representation, or inclusion of children with additional needs).

N/A

Source 1