The Learning Variability Network Exchange (LEVANTE) brings together researchers from around the world aiming to capture the richness and diversity of child development and learning.
Only by conducting open-access, cutting-edge research can we enhance our knowledge on learning and developmental variability.
Delve into the science of learning variability, explore cutting-edge research, and discover practical insights to enhance learning for all.
Home » Science » LEVANTE core assessment measures
Measure selection for a large-scale study is necessarily an iterative process; few measures meet our requirements without modification. Measures described in this site are implemented in the LEVANTE infrastructure and carried forward into the LEVANTE pilot sites. These measures are currently available in English, Spanish and German.
The age range for the LEVANTE framework is 2 – 12 years. However, selecting measures for young children ages 2 – 5 carries a number of special challenges. In particular, there is far less work on the development of direct assessments for children in this younger age group, and so it is more challenging to identify measures that have strong evidence for reliability/validity, successful internationalization, and are openly available. In light of these challenges, LEVANTE begins data collection with children ages 5 –12 and engages in iterative piloting of measures for children ages 2 – 5. In some cases, these early childhood measures will be downward extensions of measures adopted for older children, while in other cases we will adopt different measures more suitable for younger children.
The measures described below address both the general constructs and the specific tasks and measures to be adopted by the LEVANTE framework for children ages 5 – 12. In some cases these are pre-existing tasks. But in other cases, particular tasks have been instantiated with many different sets of parameters across the literature without consensus as to a single standard implementation. And even in the case of well-known measures, length constraints may make using some tasks infeasible without modification. Thus, in most cases the details of the measures have been modified to meet the specific requirements of LEVANTE, in response to both length constraints and psychometric assessment from pilot data.
Our core constructs for direct assessment were selected based on an interest in key learning outcomes and their precursors in early childhood, combined with the goal of creating a holistic assessment of individual children.
Direct assessment (60 minutes per child)
Primary caregivers provide a wealth of information about the contextual variables surrounding their children (e.g., home, school, community, family, etc.) as well as their children’s interactions within these contexts (e.g., health, well-being, emotion, etc.).
Caregiver report (60 minutes)
LEVANTE aims to characterize both the home environment of individual children but also other meaningful contexts, including their school and geographic (neighborhood) environments. We provide measures that allow sites to collect school and classroom characteristics from children’s teachers (whenever possible based on format of administration). We will also use geographic databases in combination with home geolocation data to extract a key set of geographic/environmental variables, to be de-identified locally.
Teacher and Site-level Reports
The following guiding principles were used in measure selection:
Rapid Online Assessment of Reading (ROAR)
Language and literacy task descriptions
ROAR Vocabulary. Children are presented with a word (e.g., apple) and two pictures (e.g., apple, carrot). Children must select the correct picture over the distractor picture.
ROAR Phonological Awareness. Children are presented with an object (e.g., a mouse) and must select the choice that matches on the basis of starting or ending with the same sound (e.g., a map) over two distractor choices (e.g., sun, ball).
ROAR Word Recognition. In this lexical decision task, children are presented with either a real word (e.g., “pig”) or non-word (e.g., “xop”) on a screen. Children must press one key if the stimuli is a word, and another if the stimuli is a non-word.
ROAR Sentence Reading. In this true/false decision task, children are presented with either a true sentence (e.g., “You can read books at school”) or a false sentence (e.g., “Apples are blue”) on a screen. Children must press the right key if the sentence is false, and the left key if the sentence is true.
Early Grade Mathematics Assessment (EGMA)
Number Line Estimation
Number and numeracy task descriptions (All EGMA task descriptions adapted from RTI documentation.)
EGMA Number Identification. The child hears a number read aloud and is asked to select the number from four choices.
EGMA Number Comparison. The child is presented with two numbers and asked to select the largest.
EGMA Missing Number. The child is presented a number pattern with three numbers and one blank space somewhere in the pattern. The child is asked to complete the pattern by selecting from a set of options which number goes in the blank space.
EGMA Level 1 Addition. The child is presented with several simple addition problems, asked to solve them without using pencil and paper, and told to select their answer from a set of options.
EGMA Level 2 Addition. The child is presented with several two-digit addition problems and asked to solve. They are presented with a set of options from which to select their answer.
EGMA Level 1 Subtraction. The child is presented with several simple subtraction problems asked to solve them without using pencil and paper, and told to select their answer from a set of options.
EGMA Level 2 Subtraction. The child is presented with several two-digit subtraction problems and asked to solve. They are given the option to use pencil and paper. They are presented with a set of options from which to select their answer.
EGMA Multiplication. The child is presented with single and double digit multiplication problems to solve. They are presented with a set of options from which to select their answer.
EGMA Fractions. The child is presented with addition and subtraction fraction problems to solve. They are presented with a set of options from which to select their answer.
Number Line Estimation Task. Children are shown number lines with anchor points at both ends. Children are presented with a number and asked to move a slider to mark where the number would be on the line. Alternatively, children are asked to select a number from a set of options to represent a pre-marked location on the number line (Laski & Siegler, 2007).
Matrix reasoning
Reasoning task description
Matrix Reasoning. Children are presented with a visual geometric design with a missing piece, and must select one of four choices to fill in the missing piece.
Memory Game (working memory; Vugs et al., 2017), Hearts and Flowers (response inhibition; Camerota et al., 2019; Davidson et al., 2006), Something’s The Same (set shifting; Jacques & Zelazo, 2001; Willoughby et al., 2011)
Minnesota Executive Function Scale. Our decision is to include it in the pilot round of LEVANTE data collection and compare its performance to that of the open measures.
Executive function task descriptions
Memory Game. Children see a 2 x 2 grid of squares. The squares light up in a pattern. Children watch and memorize the pattern, then select the squares in the same pattern.
Hearts and Flowers. Children learn to press a button on the same side of the screen when they see an image of a heart and learn to press a button on the opposite side of the screen when they see a flower. Task blocks are hearts-only, flowers-only, and mixed (hearts and flowers; description adapted from Camerota et al., 2019).
Something’s the Same. Children are shown two pictures that are similar along one dimension (shape, color, or size) and are told the dimension of similarity. The children then see the same two pictures plus a new third picture. The third picture is similar to one of the first two pictures along a dimension that is different from that of the first two pictures (e.g., if the first two pictures were similar along the dimension of shape, the third picture would be similar to one of the first two pictures along the dimension of color or size). Children are asked to choose which of the two original pictures is the same as the new third picture, requiring an attention shift to a new dimension of similarity (Willoughby et al., 2011).
MEFS. Children sort virtual cards into different boxes based on rules related to color and shape (Carlson & Zelazo, 2014).
Theory of Mind, Emotion Recognition, and Hostile Attribution
Social cognition task descriptions
Theory of Mind and Emotion Recognition. Children listen to brief vignettes with accompanying pictures tapping constructs of theory of mind (e.g., reasoning about desires and beliefs, moral blameworthiness, mistaken referents, second order belief-desire reasoning). Children answer multiple-choice questions about the scenarios.
Hostile Attribution. Children listen to brief vignettes with accompanying pictures describing situations involving others. Children then respond to questions asking how they perceive the actions of others (e.g., was something done “on purpose” or “by accident”).
Mental Rotation
Spatial cognition task descriptions
Mental Rotation Task. Children are presented with a picture of a target item (e.g., a simple picture of a duck). Then, they are presented with two choice items (e.g., two silhouettes of ducks). One of the choice items matches the target item when rotated; the other choice item does not match the target item when rotated. The child selects the choice item that matches the target item.
Children respond to 20 questions regarding their peer and teacher relationships, academic competence, and school experience. (questions inspired by the California Healthy Kids Survey, the Education Department School Climate Survey, the Effective School Battery, and the Self-Description Questionnaire from the Early Childhood Longitudinal Study).
Child survey topics
A primary caregiver for each participant completes a set of questions about their family, themselves, and the participant child. This person could be the mother, father, or other primary caregiver who knows the family and child well. Caregiver report constructs comprise two surveys: Demographics and General.
The Demographic Survey collects both one-time background information as well as follow up information detailing any changes or updates that may occur between data collection timepoints. Demographic information may contain questions that are both sensitive for participants and subject to regulatory control due to privacy and other restrictions. Individual sites will collect these data at the site level and be able to remove questions that they are unable to ask in their specific context. Further, demographic data will be pre-processed to ensure its compliance with anonymization standards prior to sharing with the LDR.
Demographic survey topics
General
The General Survey collects information at each data collection time point from a primary caregiver about the participant child in the following domains:
Health and well-being
Socioemotional development
Executive function and self-regulation
Home environment
Parenting
Caregiver well-being
Each participating LEVANTE data collection site will provide meta-data to be linked with their data about the specifics of their data collection site. These details will include how sampling relates to the broader population of the site location in terms of sociodemographic characteristics.
Sites will also be asked to provide other meta-data for the population from which they are recruiting including physical landscape, climate, sustenance styles, community composition, salient cultural norms around parenting and childhood, typical structure of schooling, and prevalence of technology.
When feasible, particularly in school-based data collection, children’s teachers complete a short set of questions to capture both classroom and broader school context.
To measure environmental variables, sites will use identifiable location data to derive a set of de-identified geographic features for each child’s primary household location (and in cases of school-based administration, school location). These will include two sub-constructs, neighborhood built environment measures (e.g., population density (rural/urban classification), greenspace, poverty, local area inequality (GINI coefficient), and walkability score) and general environmental measures (e.g., average temperature, temperature at date of administration, heat index, daylight, night-time light pollution, noise pollution, and air pollution metrics).