Python 分类调查的分析方法
我的调查结果如下:Python 分类调查的分析方法,python,excel,pandas,data-analysis,Python,Excel,Pandas,Data Analysis,我的调查结果如下: Q1 Q2 Q3 Very satisfied Much shorter than I expected 10 Very satisfied About what I expected 10 Very satisfied About what I expected 8 Very satisfied Much shorter than I
Q1 Q2 Q3
Very satisfied Much shorter than I expected 10
Very satisfied About what I expected 10
Very satisfied About what I expected 8
Very satisfied Much shorter than I expected 10
Satisfied About what I expected 4
Very satisfied Much shorter than I expected 10
Satisfied About what I expected 8
Satisfied Much shorter than I expected 10
Very satisfied Shorter than I expected 9
Very satisfied Much shorter than I expected 10
Satisfied Shorter than I expected 8
Satisfied About what I expected 8
Satisfied Shorter than I expected 5
Very satisfied Shorter than I expected 10
Very satisfied Much shorter than I expected 9
Very satisfied Much shorter than I expected 10
Satisfied Much shorter than I expected 9
Very satisfied About what I expected 9
Very satisfied About what I expected 10
Very satisfied Shorter than I expected 10
Very satisfied Much shorter than I expected 10
Very satisfied About what I expected 10
Neutral Shorter than I expected 8
Very satisfied Shorter than I expected 6
Satisfied About what I expected 8
Very satisfied Much shorter than I expected 10
Very satisfied Shorter than I expected 9
Unsatisfied About what I expected 3
Very satisfied Much shorter than I expected 10
Satisfied Shorter than I expected 9
Neutral Shorter than I expected 6
Unsatisfied Did not receive a response 1
Very satisfied Much shorter than I expected 10
Very unsatisfied About what I expected 1
Very satisfied Shorter than I expected 10
Very satisfied Shorter than I expected 8
回答以下问题的最佳方式是什么:如果回答者回答Q2时“比预期短得多”,那么Q3的概率是10
我正在寻找明确的答案,或者如何指导我在未来可以学会这样做。我想在excel或pandas中完成此任务
我可以使用逻辑回归并为Q2分配虚拟变量吗?我是否可以创建一个相关矩阵,以查看Q2的响应与Q3的相关程度?您可以使用以下公式进行计算:
=(COUNTIFS($B$2:$B$37,“比我预期的短得多”)、$C$2:$C$37,10)/COUNTIF($B42:$B$37,“比我预期的短得多”)*100
我们给它范围B2:B37
来检查答案“比我预期的要短得多”
当我们找到它时,我们检查他们是否给第三季度打10分
然后,我们取他们回答“比我预期的要短得多”的总次数,乘以100,四舍五入到小数点后2位
公式示例:
您还可以修改此公式以接受不同的参数:
=(COUNTIFS($B$2:$B$37,E2,$C$2:$C$37,F2)/COUNTIF($B$2:$B$37,E2))*100
如果你只是想知道计算概率的数学,以下是步骤:
定义你的事件和结果
将事件数除以可能的结果数
把答案乘以100,使之成为一个百分比
用百分比作为你的答案
…如果他们选择了“比预期短得多”,那么第三季度的概率是多少?看起来是100%,不是吗?现在还不清楚你在寻找什么,或者如何找到它。如果他们选择“关于我期望的”怎么办?第三季度
的概率是1/3…还有,到目前为止你尝试了什么?我更新了整个调查。我不认为仅仅计算出符合我问题的答案的百分比就能给出实际的概率。那么你的问题是什么?如何计算概率?我的问题如上所述“如果响应者回答第二个问题时“比预期短得多”,那么他们回答第三个问题的概率是多少?答案为“10”。我投票结束这个问题,因为这是一个关于计算概率或其他数学运算的问题,这不是一个编程问题。谢谢你的回答。不过,我不打算计算回答的百分比。我希望使用某种类型的建模或统计技术(贝叶斯、逻辑回归、独立性卡方检验等)。在这里,我们不能假设调查答案的分布代表了更大的人口。由于样本量太小,我们的统计误差会很大。@precision\u V5您实际上还没有说出您想要什么。你真的只是在问“你能教我分析数据的所有方法吗?”我不明白你在问什么。1。如果响应者回答Q2时“比预期短得多”,那么Q3为10的概率是多少正在寻求帮助回答此问题2。我将如何在excel或pandas中实现这样的解决方案-寻求帮助实现这一点以上是我的两个问题。我正试图提供尽可能多的信息,但现在看来我必须非常明确地说明这一点。这正是我已经为你们回答的问题。给定数据集,答案是83.33%
。我还向您展示了如何在excel中获得此解决方案。您知道有许多不同的方法来计算概率,对吗?我回答了你的答案,为什么我不认为它是准确的,但我要再说一次——我们不能在这里假设调查答案的分布代表了更大的人口。由于样本量很小,我们的统计误差会很大。上述说法有意义吗?正因为如此,我正在寻找一种更可靠的方法来衡量第二季度和第三季度之间的相关性。