Patterns of Observed Child Participation and Proximity to a Small Group with Teacher in Swedish Preschool Free Play

Swedish preschool is a preschool for all children and free play is an important educational activity. The question is whether all children are participating (ie, are attending important activities and are engaged when they are there) to a sufficient extent in preschool free play? This study aims to explore children's observed participation based on quantitative observational data and a person-oriented method of analysis. A series of systematic instantaneous observations á 3 seconds (M = 19.50, SD = 8.40) of individual children (n = 453) with a minimum age of 36 months, were carried out continuously for 1-2 full days in 56 preschool units in southeast Sweden. Only observations where the overall activity (activity setting) was free play (indoors and outdoors) were used in the analysis. The summed observations resulted in frequency data for four types of activities; pretend play, associative / cooperative interactions, proximity to a small group incl. teachers, and level of engagement (level of focus or effort in the activity) and were selected as important based on the Swedish preschool context and previous research. Proportions (divided by the individual's total number of observations in free play) of the individual activities were calculated. The four proportion variables were used in a series of cluster analyzes to find distinct and interpretable clusters or patterns of participation where the children in a cluster were highly similar in all variables, and at the same time were very different from children in the other clusters. The clusters were then compared based on child characteristics (age, gender, need for special support, receiving mother tongue support) answered by preschool staff, and preschool unit characteristics (number of children, number of staff, children / staff ratio, number of children with mother tongue support, number of children in need of special support, number resource staff) answered by the preschool director / principal, to see if certain characteristics were more common in some clusters than in others.

The results identified eight distinct and interpretable patterns that could be arranged from higher to lower participation. Most interesting were two clusters of children that indicated low to very low observed participation, and with high and low proximity to a small group of teachers, respectively. The children in the cluster with a very low participation pattern and low proximity to a small group of teachers were more often children with another mother tongue, and more children came from units with more resource staff. It is noteworthy that few of these children were considered to need special support according to the preschool staff. The children also came from different preschool units. Children in the cluster with a low participation pattern and high proximity to a small group of teachers more often came from preschool units where more children had a different mother tongue. No significant difference between the clusters was noted in the number of children in need of special support, although trends could be discerned.

Data was collected with systematic observations with the help of manual-based instrument Child Observation in Preschool (COP), adapted to the Swedish preschool context. The observations consists of many snapshots of individual children across a preschool day, at least. Several aspects of the child's current activity are coded. Individual data is summed across the observations, and proportions of different activities (pretend play, associative/cooperative interactions, proximity to small group with teacher, level of engagement) were calculated. Frequency and proportionate data from the observations are available in the data set, as well as child and preschool unit background information.

瑞典学前教育面向所有儿童，自由游戏是一项重要的教育活动。核心研究问题为：所有儿童是否都能充分参与学前自由游戏，即是否参与了关键活动且在参与过程中保持投入状态？本研究旨在基于量化观察数据与以人为导向的分析方法，探究儿童的学前参与情况。研究对瑞典东南部56所学前教育机构中453名（n=453）最小年龄为36个月的儿童，开展了连续1至2个完整工作日的系统性瞬时观察，每3秒记录一次（均值M=19.50，标准差SD=8.40）。分析仅纳入整体活动为自由游戏（含室内与室外场景）的观察数据。经汇总观察结果，得到四类活动的频次数据：假装游戏、联想/合作互动、靠近包含教师在内的小群体，以及活动投入程度（活动中的专注度或努力水平）。上述四类活动指标是结合瑞典学前教育场景与既往研究成果筛选出的关键维度。随后以个体在自由游戏中的总观察次数为分母，计算各项活动的参与占比。将四类占比变量纳入一系列聚类分析，以识别具有区分度且可解释的参与模式簇，使得同一簇内的儿童在所有变量上高度相似，同时与其他簇的儿童存在显著差异。后续研究基于学前教职工填报的儿童特征（年龄、性别、特殊支持需求、母语支持情况），以及学前园长填报的园所特征（儿童总数、教职工总数、师幼比、接受母语支持的儿童数量、有特殊支持需求的儿童数量、资源教职工数量），对各参与簇进行比较，以探究某些特征是否在特定簇中更为普遍。

研究结果共识别出8种具有区分度且可解释的参与模式，可按参与程度从高到低排序。最受关注的是两个呈现低至极低观察参与度的儿童簇，二者分别表现出高、低靠近教师小群体的特征。其中，参与度极低且低靠近教师小群体的簇中，儿童更常为非母语使用者，且更多来自资源教职工配置较多的学前教育机构。值得注意的是，根据学前教职工的评估，这类儿童中几乎没有被认定需要特殊支持的个体，且他们来自不同的学前教育机构。而参与度较低且高靠近教师小群体的簇，其儿童更多来自接受母语支持的儿童比例较高的学前园所。各簇在有特殊支持需求的儿童数量上未发现显著差异，但可观察到一定的趋势性特征。

本研究采用适配瑞典学前教育场景的手动观测工具《学前儿童观察量表》（Child Observation in Preschool, COP）开展系统性观察数据采集。观察内容涵盖学前一日内多名个体儿童的多帧快照式记录，对儿童当前活动的多个维度进行编码。将各观察项的个体数据汇总后，计算不同活动的参与占比，包括假装游戏、联想/合作互动、靠近教师小群体、活动投入程度。数据集包含观察所得的频次与占比数据，以及儿童与学前教育机构的背景信息。