统计学笔记

统计量

$r^2=\frac{t^2}{t^2+df}$ t检验对$H_{0}$的解释度

T-test

T -test Assumptions

1. X and Y should be random samples from two independent populations.
2. Populations are approximately normal.
3. Sample data can estimate population variances.
4. Population varances are roughly equal.

Independent Samples t-test

$Pooled- variance(合并方差)=S_{p}^2=\frac{SS_{x}+SS_{y}}{df_{x}+df_{y}}(\Sigma (x_{i}-\overline{x})^2=SS_{x})$ {样本量n不一致的时候用}

Standard Error$(S_{\overline{x}+\overline{y}})=\sqrt{\frac{s_{p}^2}{n_{x}}+\frac{s_{p}^2}{n_{y}}}$

$H_{0}:\mu_{x}-\mu_{y}=10$

$t-Statistic=\frac{\overbrace{(\overline{X}-\overline{Y})}^{Observed-difference}-\overbrace{(\mu_{x}-\mu_{y})}^{expected-difference}}{\underbrace{S_{\overline{x}-\overline{y}}}_{Standard-Error}}$

样本类型

Dependent Samples(repeated measures; Within-subject designs)

• Two conditionals
• Longitudinal
• Pre-test, post-test

Advantages

• Controls for individual differences
• Use fewer subjects
• Cost-effective
• Less time-consuming
• Less expensive

Independent Samples(Between-subject designs)

• Experimental
• Observational

Disadvantages

• Carry-over effects Second measurement can be affected byu first treatment
• Order may influence results

统计计算

$t-Statistic = \frac{"Difference-between-means"}{Standard-error}=\frac{(\overline{x_{1}}-\overline{x_{2}})-(\mu_{1}-\mu_{2})}{SE}$

$SD=\sqrt{{s_{1}^2}+{s_{2}^2}}$

$SE=\frac{s}{\sqrt{n}}=\frac{\sqrt{{s_{1}^2}+{s_{2}^2}}}{\sqrt{n}}=\sqrt{\frac{s_{1}^2}{n_{1}}+\frac{s_{2}^2}{n_{2}}}$

$df=n_{1}+n_{2}-2$