VCE Stuff > VCE Computing: Data Analytics
[Free help] Does anyone have any issues with IT Apps? I can (hopefully) help
Orson:
--- Quote from: Floatzel98 on October 24, 2015, 05:31:22 pm ---This seems like good enough question thread..
Can someone help me with understanding Question 9 from the 2012 VCAA Exam. I understand how to properly get it into 3NF but I'm at at loss trying to understand how normalizing it in each form ensures data integrity. I've looked through the examiners report and the Mark Kelly postmortems but I'm kind of getting lost in it all. Also for a question like this, do we have to show each step of the normalization or just the final form?
Any help would be great!
--- End quote ---
NF3: Fields included in tables should only be relevant to the primary key
"Fiels that aren't relevant to the primary key must be removed. Non-key fields must give information about the key and nothing but the key. In this case, that means removing "d_carRate" and "d_totalFees" from the drivers table.
This will improve efficiency as only relevant information is being processed. "d_carRate" and "d_totalFees" can be calculated in a query, and reports can be made after"
I hope this helps!
Floatzel98:
--- Quote from: Orson on October 25, 2015, 09:58:46 am ---NF3: Fields included in tables should only be relevant to the primary key
"Fiels that aren't relevant to the primary key must be removed. Non-key fields must give information about the key and nothing but the key. In this case, that means removing "d_carRate" and "d_totalFees" from the drivers table.
This will improve efficiency as only relevant information is being processed. "d_carRate" and "d_totalFees" can be calculated in a query, and reports can be made after"
I hope this helps!
--- End quote ---
Thanks Orson. What would you say about 1NF and 2NF then?
ITTeacher:
--- Quote from: Floatzel98 on October 24, 2015, 05:31:22 pm --- I'm at at loss trying to understand how normalizing it in each form ensures data integrity.
--- End quote ---
Data integrity is related to accuracy. If the data is accurate, we can trust it.
Separating lists of data enables us to more-accurately sort and filter data. (e.g. separating street address from suburb lets you filter for a particular suburb, whereas you couldn't do that before)
Including a primary key enables us to more-accurately retrieve the details of a record that may share a data value. (e.g. filtering LastName = "Smith" may retrieve multiple records, whereas filtering MemberID = "SMI0001" will retrieve the exact Smith we were looking for)
Separating tables removes redundant data, which means that the chances of updating anomalies appearing is reduced, making our data more-accurate (e.g. Member SMI0001 changes his address, if redundant data exists, the address needs to have many instances updated, whereas if there is no redundant data the address only needs to be updated once)
Moving calculated fields from the table into a query will change the calculation from being a manual calculation to an automated calculation - more-accurate (e.g. TotalFee would need to be calculated by the database operator as it's stored in the table - prone to human error. Whereas if it's calculated in a query, it can be automated using the Car Rate and Years Worked fields.)
In addition, you could also change the Years Worked field to Employment Date, which would be used to calculate Years Worked. This too would remove human error when updating, as when the data is stored in Years Worked it would require the Database Operator to change the Year for each record as the Driver completes another year of driving, a tedious task that would be prone to human error.
--- Quote from: Floatzel98 on October 24, 2015, 05:31:22 pm ---Also for a question like this, do we have to show each step of the normalization or just the final form?
--- End quote ---
Displaying your changes as a Data Structure Diagram would comprehensively show the transition from 1NF to 3NF in that you've shown the separating of fields, the primary keys & foreign keys, the relationships, the separating of tables and the removal of calculated fields.
Orson:
--- Quote from: ITTeacher on October 25, 2015, 02:01:56 pm ---Data integrity is related to accuracy. If the data is accurate, we can trust it.
--- End quote ---
Wow! Cheers mate...I learned a lot!
I just want to clarify some things:
- Doing all this normal form stuff is to increase accuracy, and decrease data duplication. This will increase accuracy as the user only needs to enter the data into the system once, and it reduces the chance of incorrect data entry significantly.
- Manipulating the system according to 1NF, 2NF and 3NF is just to get the data to its simplest form, and to always reuse data (and not dupe everything...thus more accurate yadda yadda).
Thanks mate! Could you please read my comment and let me know if what I said was correct? I copied my answer (for which I gave myself full marks). How would I improve?
Floatzel98:
Thanks heaps guys! Really clears things up.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version