2.1.2. Informed Consent

Research procedures were approved by local Ethics Committees. Full information on the study was provided in the local language before requesting consent, with frequent back-checking to ensure understanding. Participants were informed of the security procedures used to ensure confidentiality and the organizational provisions for data storage and archiving. Informed consent forms were written to be easily understood by lay persons, enabling them to understand the aims, procedures and potential risks of participation. All information was read aloud, in the presence of an adult witness, to participants who were illiterate. Respondents who decided to participate completed a written consent form, and for participants who were illiterate, a witnessed oral consent and a thumbprint in lieu of a signature was requested. All participants were informed that taking part was voluntary and that they were free to withdraw at any time. It was also made clear that taking part or not taking part would in no way affect care for their child. This consent procedure followed WHO recommendations. Once they agreed to participate, caregivers in each country were given a participant information document and were consented by different cadres of workers who had received in-depth training on the informed consent process through Good Clinical Practice (GCP). This included pediatricians, occupational therapists and physiotherapists (Brazil), nurses (Malawi) and Lady Health Workers (Pakistan). Families were recruited between 1 August 2016 and 1 December 2016.

### 2.1.3. Sampling

As this was a validation study, it was important that we sampled enough children from across the entire age range in order to adequately test all the items in each age group. In order to achieve adequate representation of children at the higher ages, we sampled children up to 42 months for analysis purposes. A stratified sampling frame was drawn up to quota sample children within eight age group strata (3-month intervals from 0 to 12 months and 6-month intervals from 13 to 42 months) by gender and from urban and rural settings. The sampling grid contained 8 × 2 × 2 = 32 cells with children to be sampled in each cell (*n* = 96 per country) (see File S1). To create subsamples for reliability and cognitive interviews, one caregiver of a child from each cell (*n =* 32 per country) was randomly selected and invited to take part in the reliability testing, and three caregivers

from different cells in each age stratum per country (*n =* 24 per country) were randomly selected for cognitive interviews.

#### *2.2. Exclusion Criteria*

The exclusion criteria were informed by existing evidence on health factors that may have affected developmental progress of children in the study. Children who had a mid-upper arm circumference (MUAC) of <12.5 [15] and infants less than six months old with a MUAC of <13 [16] at recruitment were excluded (Figure 1). Children who were unwell on the day of assessment or who had other chronic health needs (including neurodevelopmental disorders, HIV or those with recurrent infections) were excluded from the study prior to recruitment.

**Figure 1.** Flow chart demonstrating recruitment to IYCD validation study.
