In: Statistics and Probability
One way the statistical investigative cycle is taught is in four stages:
Find an example that follows this process and utilizes what we have learned about collecting data, displaying the data graphically, and making conclusions. Indicate each stage in the example.
The example that would follow above procedure can be illustrated as below:
Consider an experiment where in i wish to study the increase in gold prices till the date. The 1st step will be Collecting data, but for collect data of what? Here comes the "formulation of questions" part. We seek to study the price increase or decrease of gold, hence the formula question here will be about the gold prices at particular time periods (Month and year respectively). Therefore this will lead me to the 2nd step, Data Collection. We will collect data of gold prices at various time periods, i.e., monthly gold prices for several years, say 50 years.
Now once we have tabulated data, we now move forward to understand the collected data. Using Histogram, barplot, even regression model (Response variable=price, regressor variable=time) can be plot and line of best fit can be estimated which will give the relation between time and gold prices. Evidently, Gold prices have increased with time till now with very few fluctuations. Hence we will claim that the gold price will be increasing in future also.
To test the above claim, we will use some of statistical inference test, i.e., hypothetical testing, which is part of 3rd step, analyze data. In this we may use regression model to predict future prices of gold. we can also use histogram of bar plot to predict the future rates, but results deduced from histogram or barplot wont give statistically significant results, Hence we would ideally use Regression analysis. After this comes the 4th and last step, Interpretation of results. whatever method we used in 3rd step, will give a statistical result which will be interpreted. In case of regression analysis, the value of slope and intercept of the best fit line will be interpreted for negative or positive relation of two variable, price and time.