Study design/ sampling plan

Data for both the In-Home and In-School Surveys were gathered using an Internet panel of United States households. The Internet panel company selected was Dynata, formerly ResearchNow. Dynata is a leading supplier of on-line samples for surveys. It has large and diverse panels consisting of millions of Americans and is highly regarded as a reputable source of respondents for online surveys. Dynata utilizes appropriate industry procedures for ensuring the integrity and quality of its panels.

Surveys were designed to minimize cost, maximize data collection without the need for daily journals, rely on short-term (24–72 h) and moderate-term (30 days) recall, while maintaining an optimal survey length [16]. The surveys focused on recent recall of marker use (1–3 day recall) and monthly recall of general art material use which was included as a means to collect data in a cost-effective manner comparable to earlier survey efforts by others [13]. Art materials surveyed included acrylic paints, ceramics, chalk/pastels, craft materials, crayons, fabric paints, finger paints, glue, modeling materials, oil paints, paper mache, pencils, stamp pads, tempera paints, and watercolor paints.

Surveys were comprised of two parts: a screening survey and the main survey. Separate but similar surveys were conducted for parents and teachers. Screening surveys were completed by all persons receiving an invitation to participate. Screening survey respondents identified as parents of children under age 12, pre-school teachers, or elementary school teachers also completed the main surveys. Surveys focused on two age groups for data collection: (1) Pre-school children (also referred to as Pre-Kindergarten/ Pre-K; ages 1–4) and (2) Elementary school children (also referred to as K-6; ages 5–12).

The survey sampling plan was designed to ensure that individuals receiving invitations to participate in the study were representative of the United States population. Head of household quotas for the parent surveys were established using census information for three demographics: (1) region (Northeast, Midwest, South, and West), (2) gender (male, female) and (3) age (18 to 34, 35 to 49, 50 and older). Once a quota group (e.g., males, 18 to 34 years old, living in the Northeast) was filled, no other respondents from that group were accepted into the sample. Since no census data existed on the age, gender and regional distribution of elementary school teachers, it was assumed that teachers were geographically distributed in the same way the general population was distributed. Therefore, only regional quotas were established for teacher survey invitations. The year was divided into quarters and a nationally representative sample was drawn for each quarter shortly before it began. Samples were drawn quarterly since over longer time periods potential respondents drop out of the sample, die, cease to be qualified, etc.

Data collection (I)

In-Home Parent Survey

Participation in the main survey was limited to parents with children 1–12 years of age. Parent surveys were collected every day (approx. 2 parents per day) for one calendar year to achieve the yearly target sample size of 746 participants (186 per quarter).

Survey questions pertained to parents’ personal marker usage and marker usage for each child in the household in the designated age range (for example: number of times used, total minutes used, how long in a room where markers were being used, whether used alone or with others and use of multiple markers). Participants were also surveyed about whether the child engaged in specific art-related activities during the past 30 days.

A novel approach was used to fulfil the target sample size. Approximately two parents from different households, who had not previously participated in the survey, completed a questionnaire each day for each child in the home in the designated age range. Each parent respondent was asked about the current and previous day’s markers usage of the child(ren) for whom they were responsible. Data collected for the current day and the previous day produced a minimum of four completed surveys per day (two parents responding for a single child as today and two responding for the day as yesterday). In this way, both 24-h recall data and 48-h recall data were combined in the data collected for a single day.

(II)

In-School Teacher Survey

Participation in the main survey was limited to teachers who taught Pre-K and grades K-6. Data were collected two days per week for a calendar year, excluding Federal holidays, to achieve the intended sample size of 624 teachers (156 per quarter). Survey questions paralleled the questions used in the In-Home Study, modified for a school environment.

Teachers received questionnaires on either Tuesday or Friday of each week. Tuesday teachers completed two surveys: one with questions pertaining to the current days’ marker usage (Tuesday) and one for recall of yesterday’s marker usage (Monday). Friday teachers completed three surveys: one for the current day’s marker usage (Friday), one for yesterday’s marker usage (Thursday) and one for marker usage two days ago (Wednesday). In this way, 24, 48 and 72-h recall data were combined in the data collected for a single day.

Teacher survey participants were asked about class size, number of students who used markers, total minutes of marker usage that day and whether used alone or with others. Respondents were also asked about the use of multiple markers and whether the class engaged in other art-related activities during the previous 30 days. Due to variability in scheduled/unscheduled days off between school districts, each teacher interviewed was asked if school was in session on the days of interest.

Data analysis

Data collection, compilation, and tabulation were provided by Dynata. Tabulated and statistical values (sigma, mean, median, standard deviation, standard error, maximum value, minimum value) were computed and reported quarterly by survey question for the In-Home Parent and In-School Teacher screening and main surveys. Quarterly reports were provided in EXCEL format. Tabulated parent surveys reported results for children age 1–4, age 5–12 and parents; teacher surveys reported results for Pre-K classes, K-6 classes and teachers. All data were included in the calculations (including null data) with the exception of daily student marker use data for teachers not actively teaching.

Two-sided t tests were used to discern potential statistical differences between daily minutes of marker use data collected during the last quarter of the study (referred to as pre-pandemic and pandemic data). They were also used to determine statitistical significance between data collected for child and adult groups within each quarter (p < 0.05).