Introduction to Statistics - Types of Data

Hey readers, hope you all are doing safe and strong in COVID-19 pandemic time. Since my post on Introduction to Statistics - Measurement scales and statistical tools, here in my today's post I will describing and summarizing types of data.

What we call data in Statistics - the values of different objects collected in a survey or web services or databases, flat files, other sources such as RSS feeds or recorded values of an experiment over a time period taken together constitute what we call data in Statistics. Each value in the data is known as observation.

Classifying statistical data below,

1. based on the ways of obtaining the data
1. Primary data
2. Secondary data
2. based on the characteristic
1. Quantitative data
2. Qualitative data
3. based on the nature of the characteristic
1. Discrete data
2. Continuous data
4. based on level of measurement
1. Nominal data
2. Ordinal data
3. Interval data
4. Ratio data
5. based on time component
1. Time series data
2. Cross sectional data

Lets get a brief concept of each type of data.

Primary data

Data which are directly collected from the main source by an investigator or survey or questionnaires or agency or by anyone and these people are first to use these data. Primary data example, suppose a class teacher wants to know the mean weight of students from class eight of a particular school. If he collects data related to the weight of each students of class eight of that particular school by contacting each students personally then data so obtained by the class teacher is an example of primary data for the same class teacher.

Secondary data

Secondary data collected by an investigator or survey or questionnaires or agency or by anyone from a source which is already exists. That is, these data were originally collected by an entity or person and has been used by them at least once. And now, these data are going to be used at least second time. Secondary data example, considering the same example as discussed in case of primary data. If the class teacher collects the weight of the students from the record of that particular school, then the data thus obtained is an example of secondary data.

Note: In both the cases (primary data and secondary data) data remain the same, only way of collecting the data differs.

Quantitative data

Data are said to be quantitative data if a numerical quantity is associated with each observation. Here interval or ratio scales are used as a measurement of scale in case of quantitative data. Data based on the following characteristics generally gives quantitative type of data. Such as weight, height, ages, length, area, volume, money, temperature, humidity, size, etc. Quantitative data example, weights in kilogram of students of a class.

Qualitative data

Qualitative data is related to the quality of an object/thing, i.e. if the characteristics or attribute under study is such that it is measured only on the bases of presence/absence then the data thus obtained is known as qualitative data. Nominal and ordinal scales are generally used as a measurement of scale in case of qualitative data. For example, if a company want to do a survey for a newly launched product and if the characteristic under study is 'satisfaction' then the objects can be divided into five categories as Highly satisfied, Satisfied, Neutral, Dissatisfied, Highly dissatisfied.

Discrete data

In discrete data, if the nature of the characteristic under study is such that values of observations may be at most countable between two certain limits then corresponding data are known as discrete data. Discrete data example, number of employee present in an office in a particular day may be 80 or 150 or 500 and so on, but cannot be 80.34, 150.54, 500.67, etc. 

Continuous data

Data are said to be continuous if the measurement of the observations of a characteristic under study may be any real value between two certain limits. Continuous data example, data obtained by measuring weights of the students of a class also form continuous data because weights of students may be 42.676 kg, 39.585 kg, 45.238 kg, etc.

Nominal data

Data collected using nominal scale is called nominal data.

Ordinal data

Similarly data collected using ordinal scale is called ordinal data.

Interval data

Similarly data collected using interval scale is called interval data.

Ratio data

Similarly data collected using ratio scale is called ratio data.

For more details with examples of nominal data, ordinal data, interval data and ratio data follow this post Measurement scales and statistical tools.

Time series data

If the purpose of data collection has its connection with time then it is known as time series data. In time series data, time is one of the main variables and the data collected usually at regular interval of time related to the characteristic(s) under study show how characteristic(s) changes over the time. Time series data example, yearly expenditure of a family on different items for last three years. 

In time series data, if the purpose of the data collection has its connection with geographical location then it is known as Spatial data. For example, number of goals saved by a goalkeeper in different matches in Europa League 2021 versus different teams. 

And if the purpose of the data collection has its connection with both time and geographical location then it is known as Spacio Temporal Data. For example, data related to audience of different matches in Europa League in 2010 and 2018 will be Spacio Temporal Data.

Cross sectional data

Type of data which is collected at one point in time is known as cross sectional data. Cross sectional data example, such as income or expenditure of a family, salaries of all employees of an organization.

Summary

In this article, I reviewed the use and types of data. I also showed different examples. Thanks for reading. I hope this article helped you to understand the use and types of data. We covered classification of statistical data based on the ways of obtaining the data, based on the characteristic, based on the nature of the characteristic, based on level of measurement and based on time component.

As always, if you have a question or a suggestion related to the topic covered in this article, please add it as a comment so other readers can benefit from the discussion.

3 Common PPC Mistakes & How To Avoid Them

Not only beginners make mistakes, here I am sharing a general idea of the most common fails that experienced PPC professionals make and how to avoid them from now on.

Whether you’ve just started with PPC or work in the online ad industry for years, you feel embarrassed about a mistake that you’ve never thought you could make. After all, making mistakes is part of our human nature. But repeating mistakes is not.

Every PPC professional (beginner or expert) is guilt-ridden of making at least one mistake at some point in the ad campaign. Whether it was important or not, it’s good to identify it to make sure you’re not repeating it.

I’ve decided to focus on the most common mistakes you can make when it comes to PPC (as I personally experienced them). So here’s a checklist of the mistakes you need to avoid.

1. Wrong targeting and bidding

An important part of a successful PPC campaign is the right targeting. Just because you have the ideal demography and/or target audience in mind from previous campaigns doesn’t mean that you can guarantee future success.

The wrong audience selection for a particular campaign cannot bring the desired results, that’s why you need to be watchful when setting up your targeting options based on whatever your campaign objective is.

For example, you can limit your audience by selecting to use the option of ‘target and bid’. This option allows you to target the people who are on your retargeting list without wasting your budget on users who wouldn’t meet your criteria. However, it’s common to overlook this approach to use the ‘bid only’ option that can lead to confusing results, from lower traffic to expensive ads to lesser ROI.

Another way to make a targeting error can occur if you’re not excluding the people you don’t want to reach. If you are remarketing to a particular audience and you don’t restrict your options, you risk paying more without seeing the desired outcome. Similarly, if you choose to create multiple audience lists for your remarketing, you may risk reaching the same audience several times, which will increase the overall cost of your campaign.

Solution: Create a plan for your targeting and double-check all the options to ensure that you’re optimizing your audience as planned as possible. Keep an eye on your campaign once it starts to observe the initial results.

2. Wrong use of keywords

Most PPC professionals focus on keywords to find out new opportunities for better ROI. How often do you assess your keyword strategy though?

Commonly, some campaigns perform better than others, but it’s still useful to assess the results.

For example, you may be focusing on keywords that are too broad. This can be a good idea, but it can also a more expensive choice. You can test long-tail keywords as a more economical alternative that can lead to improved results in competitive industries.

Another common mistake is to pay no attention to the use of negative keywords. It’s easy to overlook them, but this can also affect your campaign results. Use them as part of your strategy to filter out the keywords you don’t need to avoid paying for unwanted clicks on your ads.

Last but not least, many businesses forget to bid on their own branded keywords. It may sound puzzling or unnecessary, but if your competitors bid on your brand’s keywords, then you may miss on prospects who were searching your business by using branded keywords.

Solution: Pay close attention to your keyword selections and find the ones that work better for your plans.

3. Inconsistent messaging

A profitable PPC campaign requires a great landing page. However, it’s common to design a landing page independently from the PPC ads, which leads to several possible inconsistencies:

• Design
• Copy
• Focusing on different goals
• Different personas

Your landing page should be an extension of your PPC ads. There needs to be a connection that starts with design and UX and moves to the brand, the messaging, and the KPIs.

For example, you cannot have a successful PPC campaign that focuses on increasing sales without having a landing page that doesn’t facilitate a “buy now” option.

Similarly, you cannot target a millennial without testing your landing page across all devices, especially smartphones.

These mistakes can also adversely affect your Quality Score from Google that has to do with the relevance of your ads and landing pages concerning the selected keywords. This could risk paying more to reach your target audience and it’s a mistake that you don’t want to repeat in the future.

When it comes to messaging, your PPC ads should not trick your target audience to click on something irrelevant to them. You need to create a path that is simply understandable and helps to take the next steps seem logical. It’s about applying psychographic analysis of the user to blend UX and advertising for generating the optimum results.

Solution: The next time you’re about to design a landing page, match it with your crafted PPC ads and ensure that the copy and the design are consistent and relevant to the keywords that you are competing for. Use the landing pages to guide your visitors on a journey that will bring them closer to your campaign objective.

Summing-up

The best way to avoid making common PPC mistakes is to optimize your strategy depending on your requirements. Pay close attention to your goals and adjust the targeting, the bidding, the budget, and your copy accordingly.

What’s key is to understand all the important factors that can damage your PPC campaign’s performance to ensure that you minimize the risks of any possible mistakes.