不忍淬讀基們的外行話，現科普1下統計分析的基本知識

不忍淬讀基們的外行話，現科普1下統計分析的基本知識 統計分析中呢, 所謂 vertical vs. horizontal 是純粹的外行話, 一般稍微受過點訓練的, 都知道說 time-series vs. cross-sectional. Time-series 描述的是變量在時間序列上的變化, 比如股票的價格. Cross-sectional 研究的是變量在不同 section 之間的變化, 比如不同板塊指數的相互關聯。 金融經濟學中, time-series 的應用最為廣泛。 定量經濟學中， 特別是經濟政策的研究， 經常是 time-series 同 cross-sectional 結合起來, 即所謂 panel data 分析.　統計調查呢，一般是cross-sectional ，也有用 panel data 的，比如不同收入水平的消費者在不同時期購買行為的變化等。 說信仰和道德負相關呢，是建立在統計分析的基礎上的幾項研究結果。基們如果不相信， 那就現在開始絕食禱告， 希望 Sevant 等能google 出不同的事實 … According to the Skeptic magazine vol.6 #2 1998, in multiple studies, there is a negative correlation between theism and morality. By Franzblau's 1934 study, there's a negative correlation between religiousity and honesty. Ross 1950 shows atheists and agnostics are more likely to express their willingness to help the poor than the deeply religious. 1969 Hirschi and Stark found no correlation in lawbreaking by churchgoing children and non-churchgoing children. 再幫基們1個忙， 提供Franzblau 1934年的研究文摘，具體的論文/數據，應該能在LEXIS-NEXIS上找到. 這算是偶布置給基們的作業了 (雖然偶知道. 以基們懶於思考不求甚解的態度, 真下手去找的可能性, 微乎其微 …) Franzblau, A. N. 1934. “Religious Belief and Character among Jewish Adolescents.” Teachers College Contributions to Education p. 80. Abstract: 701 students aged 12 to 16 years were given a battery of character tests, intelligence tests, and religious tests. The findings show that maturity measures (intelligence, age, etc.) are negatively related to the acceptance of religious beliefs, as are all measures of honesty and character. None of the evidence which was found supports the principle, fundamental in most religions, that acceptance of the traditional religious dogmas is creative of superior character. The "Religious Ideas Test" and the "Confession and Reporting Blank" developed for the study are given, together with the data on their validity and reliability. The bibliography lists 32 titles. [Source: PI]