## Your first Step in Data Analysis

Just as a team analyzes their opponent before the game, or an Army General examines the field before the battle, **you should know what you are facing before undertaking your Data Analysis**.

The **nature of Data can be very diverse.**

- Some data are Subjective, others are Objective, some data can be measured perfectly while others are very difficult, etc

And** you can’t use any data in any way**.

**Behind a number there is always much more than it seems at first glance**.

If you are going to develop a Data Analysis, the first thing you should do is find out which type of data you’ve got.

## Types of Data

There are 2 Types of Data:

**Quantitative Data:**- The data related to objective measures.
- Quantitative data can be:
**Continuous**or**Discrete**.

- Quantitative data can be:

- The data related to objective measures.

**Qualitative Data:**- The data related to subjective descriptions or states.
- Qualitative data can be:
**Binary**,**Nominal**or**Ordinal data**.

- Qualitative data can be:

- The data related to subjective descriptions or states.

Let’s see them in more detail:

### Quantitative Data

Quantitative Data is that **you can measure and compare Objectively**.

- Length.
- Weight.
- Sound (decibels).
- Number of people
- etc.

Depending on whether this “measurable data” can be divisible or not, we have:

**Continuous Data**.**Discrete Data**.

#### Continuous Data

Continuous Data comprises all Quantitative divisible Data.

**This Data can be infinitely divided**.

**For example**:

You can divide** a length** infinitely (maybe Planck would not agree) :

- 1 meter = 100 cm = 1,000 mm = 1,000,000 micrometers…

You can also divide** a weight** infinitely.

- 1 kilo = 1,000 grams = 1,000,000,000 nanograms…

#### Discrete Data

This Quantitative Data is that you can count, measure objectively and compare, but **it is not infivitely divisible**.

**For example**:

You can’t infinitely divide the number of** people** in a room.

- It is impossible to have 1.534272 people.

You can’t infinitely divide **objects**.

- Either you have 1 computer or you have 2, but
**you can’t have 1.5 computers**.

**Averages make people get confused.**

Many people get confused with Discrete Data because of Averages.

We all have heard that, on average, **married couples have 1.3 children** (for example).

Then, **can we divide people or not? **

- Wasn’t people a Discrete Variable?

**You can make averages of all the Data you want**.

- No matter if it is Discrete, Continuous Data, Quantitative or Qualitative.

That **doesn’t change the nature of the analyzed Data** .

Let’s see some examples:

#### Continuous and Discrete Data examples

**Continuous Data examples:**

**Weight**.

- You can infinitely divide weight (kg).

**Length**.

- You can infinitely divide length (m).

**Velocity**.

- You can have infinitely small increases in velocity (m/s, km/h, mph).

**Sound**.

- You can infinitely divide Sound volume (db).

**Temperature**.

- There are infinite degrees of Temperature (ºC or K).

**Other Physical Quantities**.

- Voltage, Intensity, Pressure, Angle, Acceleration, etc, etc.

**Discrete Data examples:**

**People**.

- You can’t infinitely divide the number of people.

**Objects: Cars, Books, Pencils, etc.**

- You can’t infinitely divide objects; there is always a minimum quantity of 1.

**Money**:

If you think carefully about it, you will realize that **it is impossible to have half a penny**.

**Money**(euros, dollars, pounds…)**is a Quantitative Discrete variable**.

Let’s now look at the other main category of Data: **Qualitative Data**.

### Qualitative Data

Unlike quantitative data,** Qualitative data is the data that serve to classify or categorize something Subjective, or describe the state of something.**

- Instead of measuring, its goal is to
**describe**something.

**You may not agree on how good a dish is** (Qualitative data) **but not on how much it weights** (Quantitative data).

**Everything can be quantified: here is the confusion**

**Everything can be quantified **in some way.

- And,
**Qualitative Data is usually quantified for commercial or Marketing purposes**.

**For example**:

- You can
**describe a person with words**. **Or**you can**develop a 100-question test that measures from 1 to 100 every aspect of his personality**.

This makes people think that all Data is Quantitative Data.

- But it is not.

Why?

Because **when you quantify certain Qualitative Data, you are quantifying Subjective Data**.

- What a person thinks about an object or other person, etc

**Quantifying some data doesn’t make that data Quantitative Data.**

- In any case, it would be Quantified data.

**For example**:

If we evaluate how handsome someone is,** what is a 9 out of 10 for you, can be a 1 out of 10 for me**.

- Nobody is right.
- It is purely subjective.

There are 3 types of Qualitative Data:

**Binary Data**.**Nominal Data**.**Ordinal Data**.

#### Binary Data

Variables (Data) that **can have only 2 possible values**.

**For example**:

**In/ Out**.**Yes/ No**.**True/ False**.- etc.

* No need to mention that, if some data had 3 possible states, this data would belong to this category.

#### Nominal Data

**Descriptive data**: Everything described with **Adjectives and Adverbs**.

**For example**:

- Colors.
**Sensations**.**Feelings**.- etc.

#### Ordinal Data

Ordinal data refers to** a Subjective Classification or a certain Order**.

This type of data is** commonly misunderstood because it is usually associated with Quantitative Data**.

**For example**:

If you tell a group of people to rank from 1 to 5 the height of some people,** they will Subjectively judge a Quantitative Data**.

**If someone is 1.80 m… Is he tall or short?**- If the observer is 1.50 m, he would be considered tall.
- On the other hand, if the observer is 2 m, he would be considered short.

**Depends completely on the Observer**.**However, nobody would dare to question that he is 180 cm tall**.- That is objective.

However, in other situations, there is no possible confusion.

**For example**:

**A race classification is no debatable**.

- If you are 2nd, you are 2nd for everybody.

Let’s see some examples of Qualitative data:

#### Binary, Nominal and Ordinal Data examples

**Binary Data examples:**

**Yes/ No**.**True/ False**.**On/ Off**.**In/ Out**.**1 / 0**.**Child/ Adult**.**National/ Foreigner**.

**Nominal Data examples:**

**Colors**.

- Blue, Yellow, Orange, etc…

**Sensations**.

- Scary, Funny, Boring, Exciting, etc.

**Flavors**.

- Acid, Sweet, Sour, etc.

**Ordinal Data examples:**

**Similarity**.

- Evaluate from 1 to 4 how similar X is to Y.

**Likability**.

- From 1 to 10, what you like most?

**Subjective size**.

- From 1 to 10, is he tall or short?

**Subjective weight**.

- Evaluate how heavy this object is.

**Ranking**:

- What position did he end up in?

Now, you may be asking yourself: “but why is this important?”.

## Use Useful and Quantifiable Data

If you know which type of data you have, you can develop a much better analysis.

You will** give data the importance it deserves**.

- If you are designing a new comfortable chair, you’ll use :
**Objective**data for calculating average weights.**Subjective**data for assessing whether people liked it or not.

**And more important**: You will be able to** Quantify Data properly and make it useful **for your analysis.

How?

As we mentioned earlier, everything can be Quantified (even colors can be classified according to their wavelength).

- But remember, that doesn’t make all variables to be Quantitative data.

When developing any Data Analysis,** you have to evaluate the variables you’ll handle and Quantify them in the most useful way for you**.

The best Quantification possible, always depends on the variable.

However, we’ll give you **3 Tips for handling your Data properly**:

**1. Handle continuous variables whenever you can**.

**2. Group non-quantifiable variables and study them independently**.

**For example**: Don’t mix Children and Adults; study them separately.

**3. Don’t mix Subjective and Objective variables**.

- It can cause confusion and misunderstandings.

Let’s see a Real example:

#### Dates - Example of Quantified Data

Imagine that** you are studying how some variable evolves during a whole year.**

**You could use dates** for developing this study.

…. Or** you could number the days from 1 to 365**.

- This is a much better option.

Think about how tricky a date-base study would be:

- It would be a mess in Excel.
- If you wanted to study sub-periods of time;weeks, 5-days… It would be a chaos.
- Each month is different in number of days.
- It is not very intuitive.

**For example:** **What feels more intuitive?**

- From 01/01/2020 to 03/03/2020 the sales increased a 20%.
- Sales increased a 20% the
**first 62 days of the year**.- And immediately, you can calculate:
**62/365 = 17% of the year**.

- And immediately, you can calculate:

If you handled days instead of dates,** you would develop a much more efficient analysis**.

By picking** the right change of variable** you’ll be able to develop much better analysis.

And to do so, you must know what kind of data you’ve got in your hands.

### Summarizing

You should always know which type of data you’ve got in your hands before developing any Data Analysis.

There are 2 Types of Data:

**Quantitative Data:**data related to objective measures.- This data can be:
**Continuous**: data that can be infinitely divided.**Discrete:**data that can’t be infinitely divided.

- This data can be:

**Qualitative Data:**data related to subjective descriptions or states.- This data can be:
**Binary:**data that can only have 2 values.**Nominal:**descriptive non-numerical data.**Ordinal data:**subjective classification or a certain order.

- This data can be: