Building Statistical models

A statistical model is way to compute data and understand how are things related to

make assumptions. One of the simplest statical models used is mean. Mean is a central

measure of tendency. Now mean is a compression of an entire data, it is an average of the

group data that we have. Although the mean is a simple model, its usefulness depends on how

well it fits the data — meaning how closely the individual data points cluster around it.

Example of mean- 2,3,4,5,6,7

Mean = 2+3+4+5+6+7/6

=27/4 = 4.5

Our mean is affected by the extreme values the extreme values in our data is also known as

the outliers. The mean summarizes the entire data.

A “fit” is nothing but a degree to which our statistical models is representing our data that has

been collected. A good fit is that is a replica of the reality. A moderate fit is somewhat

replicable of the reality there would be some differences between the model and the data but

also with great similarities, whereas, a poor fit is the data to be completely inaccurate and not

matching the model. In good fit we can have complete confidence about our predictions, in

moderate fit we can have somewhat confidence and in poor fit we cannot have confidence. It

is important to have fit statistical model and a good fit statistical data so that our predictions

(assumptions) can be well.

Population and samples

A population refers to the entire group of data/ individuals or data that we are

interested in studying. Example – a researcher conducting study for students with height

above 5’8 in all the universities in Mumbai so all the students in universities in Mumbai will
be above 5’8 will be our population. A sample is the subset of the population, a

representative of the population. Ex- a researcher conducting study for students with height

above 5’8 male age 22 in the universities in Mumbai so all the male students with height

above 5’8 male age 22 in universities Mumbai will be our sample size.

The larger the sample is the better representation of the population. We can never have access

to our entire population. So, we use samples. However, we can take separate samples of the

entire populations, there a chance that then the samples among each other can slightly differ.

That is known as the error. Error = sum of (total – mean). we need to avoid the problem of

which direction the error is in and one mathematical way to do this is to square each error that

is multiply each error by itself. So, rather than calculating the sum of errors, we calculate the

sum of squared errors. The sum of squared errors (SS) is a good measure of the accuracy of

our model. However, it is fairly obvious that the sum of squared errors is dependent upon the

amount of data that has been collected – the more data points, the higher the SS. To overcome

this problem we calculate the average error by dividing the SS by

the number of observations (N). This is the formula for Variance.

Variance is how spread out the values in a dataset are. It shows

how much the data deviates from the mean on average. And standard deviation is the spread

of data. SD is simply the square root of the variance.

. The lesser the standard deviation the closer it is to the mean. Larger sd

is a poor fit. And larger sd gives us a flatter curve in the curve

distribution. Small SD is small spread and a good fit. EXAMPLE- first lecturer had a

standard deviation of 0.55 (relatively small compared to the mean). Mean= 0.55, second

lecturer, however, had a standard deviation of 1.82 (relatively high compared to the mean).

clearly more spread from the mean, therefore this mean is a poor fit to the data.
STANDARD ERROR

A population refers to the entire group of data/ individuals or data that we are interested in

studying. A sample is the subset of the population, a representative of the population. The

larger the sample is the better representation of the population. When we take several samples

from a population there is a chance that the samples might slightly differ from each other.

This is where we use standard error. When someone takes a sample from a population, they

are taking one of many possible samples. If we were to take several samples from the same

population, then each sample has its own mean, and some of these sample means will be

different. For each of these will be a sample mean. Each sample mean deviating from

population mean is known as standard error.

Since the samples will vary from each other because they contain different members or

values of the population is known as sample variation. We can Plot this sample mean on a

frequency distribution or histogram. A sampling distribution is simply the frequency

distribution of sample means from the same population. The sampling distribution tells us

about the behaviour of samples from the population, and we’ll notice that it is centred at the

same value as the mean of the population. This means that if we took the average of all

sample means we’d get the value of the population mean. if the average of the sample means

is the same value as the population mean, then if we knew the accuracy of that average we’d

know something about how likely it is that a given sample is representative of the population.

And standard deviation is the spread of data. Small standard deviations represent a scenario

in which most data points are close to the mean, a large standard deviation represents a

situation in which data points are widely spread from the mean. The standard deviation of

sample means is known as standard error of the mean SE. Calculation of standard error

could be done by taking the difference between each sample mean and the overall mean,

squaring these differences, adding them up, and then dividing by the number of samples.
Finally, the square root of this value would need to be taken to get the standard deviation of

sample means, the standard error. When the sample is relatively small

(fewer than 30) the sampling distribution has a different shape, known as a

t-distribution.

To conclude, the standard error is the standard deviation of sample means. As such, it is a

measure of how representative a sample is likely to be of the population. A large standard

error (relative to the sample mean) means that there is a lot of variability between the means

of different samples and so the sample we have might not be representative of the population.

A small standard error indicates that most sample means are similar to the population mean

and so our sample is likely to be an accurate reflection of the population