第三周

 

Weekly Repot- No.3

 

We spent this week learning about SPSS, a software designed mainly for data analysis. In fact, as the Professor mentioned, SPSS was a software easy to manipulate and was of great use in fields regarding to mathematics and statistics. Thus, knowing how to deal with SPSS was significant and beneficial for our further exploration in these field.

 

我们本周花了一些时间学习SPSS，这是一个主要用于数据分析的软件。事实上，正如教授所提到的，SPSS是一个易于操作的软件，在数学和统计领域具有很大的应用价值。因此，了解如何处理SPSS对我们在这些领域的进一步探索具有重要意义。

 

For our first meeting, we started from getting used to the software. Unlike with Excel, I and my classmate had no experience working with SPSS, thus it really took some time for us to know how to import data directly from other source to SPSS, what the meaning of scale and nominal are and how to distinguish them, how to set up new variables and how to define their properties. After this, we learned a popular approach of data analysis called factor analysis. To understand this seemingly complex method, the Professor showed us two short videos talking digging into the mathematics rules behind the method. Through these learning sources, we began to know the meaning of the approach. Also, a general view of how the approach was developed through series of mathematical computation using matrix transformation started to take shape in our mind. Upon this, the Professor then showed us how to apply the method. He picked an example from his formerly-published paper which involved factor analysis to demonstrate how factor analysis was used in a specific case. Then came the time for the SPSS function- Dimension Reduction. We used this function to reduce several variables to fixed-number dimensions decided by calculating and comparing the eigenvalues of given data. After that, we could project all the existing samples onto our new dimensions. By observing the final plot with dimensions as axis, samples as points and factors as arrows, it was possible for us to analyze the given data in a more concise and direct way.

 

对于我们的第一次会议，我们从习惯软件开始。与Excel不同，我和我的同学没有使用SPSS的经验，因此我们真的花了一些时间才知道如何将数据直接从其他来源导入到SPSS，规模和名义的含义以及如何区分它们，如何设置新变量以及如何定义其属性。在此之后，我们学会了一种流行的数据分析方法，称为因子分析。为了理解这种看似复杂的方法，教授向我们展示了两个简短的视频，他们正在深入探讨该方法背后的数学规则。通过这些学习资源，我们开始了解该方法的含义。此外，关于如何通过使用矩阵变换的一系列数学计算开发该方法的一般视图开始在我们的脑海中形成。在此之后，教授向我们展示了如何应用该方法。他从他之前发表的论文中选择了一个例子，该论文涉及因子分析，以证明在特定情况下如何使用因子分析。然后是时间进行SPSS功能- 降维。我们使用此函数将几个变量减少到通过计算和比较给定数据的特征值而确定的固定数量维度。之后，我们可以将所有现有样本投影到新维度上。通过观察以尺寸为轴的最终图，以箭头的形式作为点和因子进行采样，我们可以更简洁直接地分析给定数据。

 

We explored two other SPSS function—— Cluster Analysis and Compare Means. We learned three kinds of Cluster Analysis: TwoStep Cluster Analysis/ K-Means Cluster Analysis/ Hierarchical Cluster Analysis. The difference between those three methods were quite obvious: TwoStep generates clusters automatically; K-Means asks for number of clusters the user want and begins with random classification, which would then be modified to give the final classification through further analysis; Hierarchical (used most often) works by calculating and comparing the distance of given samples and forms clusters of gradually-increasing size. Cluster Analysis offered us a way to gain a feeling of how the given data distributed and the closeness of different clusters. Compare Means was an approach helping calculating average, mean value, standard deviation……. It was much more convenient than Excel since SPSS could calculate all values at the same time by several additional clicks while Excel required for more manual work to input the function names one by one.

 

我们探索了另外两个SPSS函数- 聚类分析和比较均值。我们学习了三种聚类分析：两步聚类分析/ K均值聚类分析/分层聚类分析。这三种方法之间的区别非常明显：TwoStep自动生成集群；K-Means询问用户想要的簇数，并从随机分类开始，然后通过进一步分析对其进行修改以给出最终分类;分层（最常用）通过计算和比较给定样本的距离并形成逐渐增大的大小的簇来工作。聚类分析为我们提供了一种方式来了解给定数据的分布方式以及不同聚类的紧密程度。比较均值是一种帮助计算平均值，平均值，标准差......的方法。它比Excel更方便，因为SPSS可以通过几次额外的点击同时计算所有值，而Excel需要更多的手动工作来逐个输入函数名称。

 

The last meeting was used for presentation and demonstration. To test our understanding of newly-learned knowledge, the Professor left us with some data regarding to an environment problem and asked us to apply all the methods we learned to analyze the data. The presentation proceeded fluently in general due to our good preparation. There were several small typos regarding to our understanding of some of the methods and the Professor was very patient and willing to help us get through those problems.

 

上次会议用于演示和演示。为了测试我们对新学到的知识的理解，教授给我们留下了一些关于环境问题的数据，并要求我们应用我们学到的所有方法来分析数据。由于我们的良好准备，演示文稿一般流利。关于我们对某些方法的理解存在一些小错字，教授非常耐心并愿意帮助我们解决这些问题。

 

All in all, we have learned something very useful in this week through a combination of understanding and application. The Professor was very warm-heated and knew the learning materials well.

 

总而言之，通过理解和应用的结合，我们在本周学到了非常有用的东西。教授非常热情，很了解学习材料。