Shulou(Shulou.com)06/01 Report--

What this article shares to you is about the slope change model regression operation of Eviews virtual variables. The editor thinks it is very practical, so I share it with you to learn. I hope you can get something after reading this article. Let's take a look at it.

Virtual variable (Dummy Variables), also known as fictitious variable, nominal variable or dumb variable, is an artificial variable that reflects the attribute of quality. it is a quantized independent variable, usually with a value of 0 or 1. The introduction of dumb variables can make the linear regression model more complex, but the description of the problem is more concise, one equation can achieve the function of two equations, and close to the reality. This paper first introduces the regression and analysis of virtual variables including intercept change and slope change in the model.

Statistical data of per capita income of Urban residents and ownership of Color TV sets per 100 households in 1998 to establish the demand function of Color TV sets for Urban residents in China

Note: d and XD are the variables generated later.

Correlation graph analysis

Type the command:

SCAT X Y, the correlation between per capita income and the number of color TV sets is shown in the figure.

two

Define virtual variables

As can be seen from the correlation chart, there is a great difference in the ownership of the first three sample points (that is, low-income families) and the last five sample sites (middle and high income). Therefore, in order to reflect the influence of the qualitative factor of "income level", the virtual variables are set as follows:

Construct virtual variable

Method 1: enter directly using the DATA command

Method 2: define directly using the SMPL and GENR commands.

DATA D1

GENR XD=X*D1

three

Estimation model

Estimate the virtual variable model:

LS Y C X D1 XD

Then the way of introducing the virtual variable is judged by the test value, and the demand function of all kinds of families is written.

According to the above steps, the estimated result of the virtual variable model is shown in the figure.

The test of the regression coefficient of the virtual variable is significant, and the goodness of fit of the model is very high, which shows that there are obvious differences in intercept and slope between the low-income and high-income families of urban residents in China. Therefore, it is reasonable to introduce virtual variables by addition and multiplication. The demand functions of low-income families and middle-and high-income families are:

Low-income families:

Middle and high income families:

Thus it can be seen that the characteristics of the consumption demand of color TV sets of urban households in China at the present stage: for low-income families with a per capita annual income of less than 3300 yuan, the demand increases rapidly with the increase of income level, and for every 1000 yuan increase in per capita annual income, the number of 100 households will increase by 12 sets on average. For middle-and high-income families with a per capita annual income of more than 4100 yuan, although the demand is also increasing with the increase of the income level, the growth rate slows down. for every 1000 yuan increase in per capita annual income, the number of 100 households only increases by an average of 3 units.

The above is what the slope change model regression operation of Eviews virtual variables is, and the editor believes that there are some knowledge points that we may see or use in our daily work. I hope you can learn more from this article. For more details, please follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.