本研究探討2012年至2022年52個國家參與國際學生能力評量計畫的家庭經濟社會文化地位對數學素養影響的效果值、效果值成長類型、不同年度效果值差異。此效果值愈大代表教育愈不公平。研究顯示：一、2012年至2022年52個國家家庭經濟社會文化地位對數學素養影響屬於中度效果，效果值成長呈U形趨勢；2012年至2018年逐年下降，2018年到谷底，2018年之後至2022年上升；二、2012年效果值明顯高於2015年及2018年；2015年及2018年明顯低於2022年；2022年有較高的效果值，代表家庭背景對數學素養的解釋變異量大，這是受到疫情的影響，學生需要家庭更多學習資源支持，使得家庭經濟社會文化地位對學習素養的重要性提高，拉開貧富家庭對子女協助需求差距；三、2012年至2022年52個國家的家庭經濟社會文化地位對數學素養具有影響效果，但各國效果值不一；臺灣屬中度效果，解釋變異量在12.96%至18.49%之間；四、2012年至2022年歐洲、美洲、亞洲的家庭經濟社會文化地位對數學素養都屬於中度效果，但亞洲國家的家庭背景對數學素養的解釋變異量較小，歐洲國家的家庭背景對數學素養的解釋變異量較大。本研究長期觀察52個國家的教育公平，找出國際間及各國教育公平趨勢，各國應依其教育不公平程度提出改善，以促進教育公平。

This study explored the effect size, effect size growth type and effect size differences of family economic, social and cultural status (ESCS) on mathematical literacy in 52 countries from 2012 to 2022. The larger the effect size, the more inequity in terms of education. Statistical analysis of data collected from the PISA showed that: (1) From 2012 to 2022, the impact of family ESCS on literacy in 52 countries was medium in effect size, and the effect size growth showed U-shaped trend, declining year by year from 2012 to 2018, reaching a zenith in 2018, and rising from 2018 to 2022; (2) The effect size in 2012 was significantly higher than that in 2015 and 2018, with 2015 and 2018 being significantly lower than 2022, with the higher effect size in 2022 indicating that family background explains a large amount of the variance observed in mathematical literacy. This was due to the impact of the COVID-19 epidemic. Students need more learning support from their families, which made family ESCS more important for mathematical literacy, and widened the gap in the need for assistance between children from rich and poor families; (3) From 2012 to 2022, the effect sizes of family ESCS on mathematical literacy in 52 countries were significant, but the effect sizes varied from country to country, with Taiwan having a medium effect size, with the explained variance ranging from 12.96% to 18.49%; and (4) From 2012 to 2022, the effect sizes of family ESCS on mathematical literacy in European, American, and Asian countries were significant, all of which were medium effect sizes, although Asian countries’ family background explained the low amount of variance in mathematical literacy; a phenomenon that was not seen European countries. The most important feature of this study is that it longitudinally observed educational equity, identifying trends both internationally and within countries, and suggested which countries should propose improvements to promote a more equitable environment, based on the degree of inequality observed.