MHB What independence test should I use for this group of data? <-----------

AI Thread Summary
For analyzing the relationship between crime rates and unemployment rates in cities, the chi-squared test is suggested as an appropriate method for testing independence. The discussion emphasizes the importance of using the correct statistical test rather than focusing on correlation coefficients. A method of least squares is mentioned as a technique for approximating relationships in the data, particularly when fitting a linear model. The user is encouraged to calculate the line of least squares to visualize the relationship effectively. Overall, the chi-squared test is recommended for the user's specific data analysis needs.
hapx
Messages
1
Reaction score
0
Hey Guys! I am working on a math project and I am stumped. I'm not sure weather I should use the chi-squared test or another test with my set of data. I am testing weather there is a relationship between crime rates and the unemployment rate of cities. Could someone please help? I'm not testing for the "r" correlation coefficient, but for the independence. I've attached my data. View attachment 3781 .
Please help me, I would really appreciate it! :-)
 

Attachments

  • crime ratio.PNG
    crime ratio.PNG
    16.4 KB · Views: 100
Last edited:
Physics news on Phys.org
Welcome on MHB hapx!...

we introduce a simple algorithm that will be very useful in the examination of the data collected. Suppose we have n + 1 data formed in n + 1 points $x_{k},\ k = 0,1, ..., n$. Typically this is the situation that occurs when trying to characterize a phenomenon 'experimental'.Suppose we want to approximate the function 'unknown' y (x) for which we know in n + 1 samples detected 'experimentally' [therefore subjectto 'experimental errors' ...] with a polynomial of degree m with m<n so that it is …$\displaystyle y(x) \sim a_{0} + a_{1}\ x + a_{2}\ x^{2} + … + a_{m}\ x^{m} =p(x)\ (1)$ An approach known as the 'method of least squares' evaluates the coefficients $a_{i},\ i = 0,1, ..., m$ imposing to be minimum the quantity …$\displaystyle S= \sum_{k=0}^{n} [y_{k} – p(x_{k})]^{2}\ (2)$A case particularly 'simple' to deal is when the 'polynomial least squares' p (x) has degree m = 1, that is a straight line which iscalled 'line of least squares'. Since it is $p(x) = a_{0} + a{1}\ ⋅x$, is to determine only two unknowns: $a_{0}$ and $a_{1}$.With the usual techniques of the analysis is that the values of the two unknowns that minimize the expression (2) are ... $\displaystyle a_{0}= \frac{s_{2}\ t_{0} – s_{1}\ t_{1}}{s_{0}\ s_{2} –s_{1}^{2}}$ $\displaystyle a_{1}= \frac{s_{0}\ t_{1} – s_{1}\ t_{0}}{s_{0}\ s_{2} –s_{1}^{2}}\ (3)$… being...$\displaystyle s_{i}= \sum_{k=0}^{n} (x_{k})^{i}$$\displaystyle t_{i}= \sum_{k=0}^{n} y_{k}\ (x_{k})^{i},\ i=0,1,2\ (4)$

In Your case is n = 19, so You have all the data You need ... once You have calculated the line of least squares to plot everything you can in the manner shown in the picture ...

i99363989._szw1280h1280_.jpg
In this case, the linear relationship between the data appears to be very good ...

Kind regards

$\chi$ $\sigma$
 
Back
Top