07 Evaluate: Unit 02 Lesson 03

Choose the techniques that improve the collection and processing of a large data set.

 

biasing the data to reduce its size

 

using multiple computers to process it and multiple servers to store it - ANSWER

 

removing data that seems unnecessary

 

cleaning the data to make it smaller

• Having enough computer power- If the data set is very large, you may need more than a single computer to contain the data and to process it. Some data sets must be stored across several disk servers  or server farms because of the large amount of storage space needed and processing the data may need to be spread to several parallel computer systems because it would take a single computer too long to process.

Choose the statement that is most accurate.

 

even data that has uniform format must be cleaned

 

if you lack powerful data tools then you must collect even more data to make up for it

 

collecting more data will solve a data bias

 

getting good information from data requires powerful tools and the skills to use them - ANSWER

Having the right data tools- If you are going to input homework grades for a class of 30 students to try to look for trends and averages, you might be able to use paper and pencil, but if you are going to track a few thousand students' grades, you will need a spreadsheet program or maybe an even more powerful, sophisticated data management tool. Depending on what you want to do with the data, some software tools are better than others. Some tools are good at making various charts, other can be used to make projections of trends that the data shows. The bottom line is that users need the proper tools and the skill to use those tools in order to extract information from data. Without powerful software tools and the skill to use them, collecting data will not yield valuable information.

Choose the option that describes what you should do if the format of the data is not uniform. For example, some data uses number dates like "03/24/2025", other data uses written dates like "March 24, 2025"?

 

you should clean the data - ANSWER

 

you should get more data

 

you should reject the data

 

you should bias the data

The data format is not uniform- Depending on how it was collected, the same type of data may look different, such as capitalization, abbreviations, feet vs. meters, and many alternate ways to record the same information. When data is dissimilar in this way it is difficult to compare and it must be "cleaned", meaning that it will be converted to the same format without changing the meaning of the data. An example of cleaning data would be changing "TX" to "Texas" to make all state references the same.

Choose the best description of data "bias".

 

when people don't like what the data shows

 

when there is too much data

 

when the data was skewed because of the way it was collected - ANSWER

 

when the data indicates a particular trend

• Bias in the data- Information can be skewed depending on the way it is collected. This is something that needs to be investigated ahead of time and can possibly be corrected by using different gathering techniques. Navy ships used to take ocean water temperatures all around the globe using buckets of seawater. It turned out that the readings were actually showing too cool because they used canvas buckets which were being cooled by evaporation. Taking more of the same samples would not change this. They finally began using different means to get the water. 

Choose the best description of data "cleaning".

 

changing the meaning of the data without changing the format

 

making sure there is no offensive data

 

removing data that does not agree with the rest of the data

 

changing the formats to match without changing the meaning of the data - ANSWER

Depending on how it was collected, the same type of data may look different, such as capitalization, abbreviations, feet vs. meters, and many alternate ways to record the same information. When data is dissimilar in this way it is difficult to compare and it must be "cleaned", meaning that it will be converted to the same format without changing the meaning of the data. An example of cleaning data would be changing "TX" to "Texas" to make all state references the same.

Choose the description of where metadata is found.

 

metadata is not found anywhere, it is computed using an algorithm

 

often in the data file itself - ANSWER

 

never in the data file

 

it is extracted from the data

Where is metadata found? There are many places, but the list below describes the most common

• embedded metadata- This is stored in the file itself. Many file types have a "header" area so that the metadata can be easily located without searching through the actual data contents of the file. Image files can have information such as camera settings, image dimensions, even GPS location. Document files can contain metadata such as author name, creation and modification dates.

• sidecar files- These are metadata files that are paired with the data file. 

• databases- This metadata is separate from the data file. An example of this could be a company saving information about its file activities, including which users sent which files, at what time, how much network bandwidth was used and other transmission related information.

Choose the best example of metadata.

 

when you detect a trend in the data

 

the words typed in a text

 

the time a text was sent - ANSWER

 

a phone conversation

What is metadata? It is data about the data. For example, if you send texts from your phone, The content of the texts themselves would be considered data, but the associated metadata would be information like:

• The time and data the texts were sent

• The size of the texts

• Who the texts were sent to

• Any other information that might be part of the transmission, such as GPS location, cell towers contacted, etc.

Choose the best description of metadata.

 

it is data about data - ANSWER

 

it is what you extract from information

 

it is like data only larger

 

it is like data only smaller

What is metadata? It is data about the data. For example, if you send texts from your phone, The content of the texts themselves would be considered data, but the associated metadata would be information like:

• The time and data the texts were sent

• The size of the texts

• Who the texts were sent to

• Any other information that might be part of the transmission, such as GPS location, cell towers contacted, etc.

Choose the best description of "causal relationship".

 

being just friends but not serious

 

when one event causes another event - ANSWER

 

the same as "casual relationship"

 

when one event does not really cause the other event

What is a causal relationship? Be careful that you don't misread the word "causal" for the word "casual". The word "causal" means that one event causes another to happen, like rain causes the ground to be wet. The word "casual" means relaxed, less formal, like dressing casually. As we analyze data, when we observe two things together, it may seem that one thing caused the other. For example, in an analysis of voter data, you may find that rural areas tend to have more republicans. Is there a causal relationship? Did living in the country cause people to become republican, or do republicans tend to move to the country? Without more information it may be mistaken to assume a causal relationship.

Choose the best description of the difference between data and information.

 

data is extracted from information

 

information becomes data after it is processed

 

data and information are the same thing

 

information is extracted from data - ANSWER

What is data? For our purposes, let's say that "data" is the raw facts that may or may not be well organized.

What is information? Information is what gets extracted from raw data. It may include facts, patterns, and trends.