Building Your Data Analysis Toolbox: Understanding Descriptive and Inferential Statistics
As a small business owner, you make important decisions every day. To make the best decisions, however, you need reliable information. That’s where statistics come in. Statistics are powerful tools to analyze data and draw conclusions. There are two major types of statistics: descriptive and inferential.
Descriptive Statistics
Descriptive statistics summarize data and help you understand a specific set of observations. The most common tools used in descriptive statistics include the measures of center and measures of variance.
Measures of center give you one number that best represents the data. Examples of these measures are the mean, calculated by adding up all the values and dividing by the number of values, and the median, the middle number when all numbers are ranked according to value. The median is good to use when you have values that are very different than the rest of your data since it isn’t affected by these values as much as the mean is.
Measures of variance tell you how far apart the data is and include the range and standard deviation. The range is a simple calculation done by subtracting the lowest value from the highest value. The standard deviation, on the other hand, is more complex to calculate, but gives a better understanding of how close to the center point your values lie and isn’t as affected by a single value that’s abnormally high or low.
Suppose you run a small retail store and you want to know if you have more people visiting your store on weekdays or on weekends so that you can schedule the right number of employees accordingly. You could get the mean or the median number of customers on weekdays and weekends to understand which days you need to staff more employees. You could also use the range or standard deviation to understand how the number of visitors on those days fluctuates. Maybe your analysis shows that weekends have the most customers visiting your store, but that number also fluctuates a lot depending on when there are sports games nearby. This means that you need to consider other factors when deciding how many people to staff on the weekends and not just what day of the week it is. The weekdays may not have as many visitors, but they are much more consistent and therefore easier to staff.
Inferential Statistics
What about when it isn’t feasible to gather data from the entire population? Inferential statistics are used in this situation by taking data from a smaller sample and using that to infer characteristics about the larger population. This process is done by using statistical models.
The way that the sample is gathered and its size are also both very important things to consider when you want to have confidence in inferential statistics. Random sampling is important here to ensure that there aren’t any other factors that could explain the patterns you see. You also need to make sure that your sample is large enough to capture more of the variability of the population. You’re definitely more confident that you’re getting something good with that Amazon purchase if there are 20,000 reviews instead of just 2.
Let’s say that you want to sell a new kind of cookie near the checkout of your retail store and want to know which one customers would prefer most. It definitely wouldn’t make sense to ask every person in the area around your store to try each cookie and tell you which one they would be more likely to buy. Here, you could randomly ask customers when they check out to try two different cookies and tell you which one they like better. Once you get a large enough number of people that told you which cookie they like better, you can reasonably guess that the customers who didn’t try the two cookies will like the winning cookie better as well.
Conclusion
In conclusion, understanding the difference between descriptive and inferential statistics is crucial for making informed decisions as a small business owner. Descriptive statistics summarize data and help you understand a specific set of observations, while inferential statistics use sample data to infer characteristics about a larger population. Both types of statistics are powerful for making decisions, but it’s essential to use the right type of statistics for the situation.
