2024-02-27

Point Estimation

Point estimation takes place when sample data is taken to form a statistic, in which can then be used as an estimate for an unknown parameter.

Examples of Sample Data

  • Sample variance (\(s^2\)) can estimate for population variance (\(\sigma^2\))
  • Sample mean (\(\bar{x}\)) can estimate for population mean (\(\mu\))

Mathematical Expression of Sample mean

Sample Mean = \(\bar{x}\) Population Mean = \(\mu\) Sample Size = n Individual Sample Value = x_i \[ \bar{x}=\frac{1}{n}\sum_{i=1}^{n}x_i \]

Limit Theorem

Central Limit Theorem: For a substantially large sample size the distribution can be approximated as normal, and with this it can be concluded… \[ \mu_{\bar{x}}=\mu \] and \[ \sigma_{\bar{x}}=\frac{\sigma}{\sqrt{n}} \]

Consider this..

Weights of weiner dogs in the entire US would be a hard task to calculate by hand, but what we can do instead is create a large enough sample data set (weights contained in 1 state) that it can then be used to approxiamte the weight of any other weiner dog in the country

Example of a Data Set Shown on a Scatter Plot

## Loading required package: ggplot2
## 
## Attaching package: 'plotly'
## The following object is masked from 'package:ggplot2':
## 
##     last_plot
## The following object is masked from 'package:stats':
## 
##     filter
## The following object is masked from 'package:graphics':
## 
##     layout

The code behind the scenes

Since the example was theoretical, and unfortunately the data of weiner dogs weight for Nebraska has not been released, a random data set needed to be produced first to create the scatter plot.

library(plotly)
set.seed(123)
x<- rnorm(200,10,5)
y<- rnorm(200,10,5)

fig<- plot_ly(x=~x,y=~y, type = 'scatter',mode='markers')

fig<- fig%>% layout(title= 'Weights of Weiner Dogs in Nebraska',
    xaxis=list(title = 'Age'),
    yaxis = list(title = 'Weight'))

To produce the graph a final code line of the named plot is needed (“fig”)

Graph with Line of Best Fit

## `geom_smooth()` using formula = 'y ~ x'

Second Graph with Line of Best Fit

## `geom_smooth()` using formula = 'y ~ x'

Conclusion to point estimation

In general, though they may not present the direct mean, graphs provide a great visual to how spiratic the data sets might be. With larger sets of sample data, one can conclude a sample mean (or variance depending on what is presented graphically) and with this statistic found, an approximated parameter for the entire population can be made.