Welcome, Guest. Please login or register.

October 24, 2019, 12:30:28 am

Author Topic: Standard Deviation  (Read 364 times)  Share 

0 Members and 1 Guest are viewing this topic.

Zaljc

  • Forum Regular
  • **
  • Posts: 78
  • Respect: +7
Standard Deviation
« on: October 29, 2018, 05:51:04 pm »
0
Hey guys,

Im pretty sure that SD allows us to see the spread of data and how far away everything is from the mean. But I was wondering if a smaller SD means our results are more likely to be reproducible.

The reason why I think that- because SD is how far data is away from the mean. So if we have a smaller SD, that means there is less variation and all the data is closer to the mean. And shouldn’t that mean that we are more likely to reproduce the same results?

Cheers, Zoe
2017- Biology

2018- Chemistry, English, Methods, Physics & Business Managment

KiNSKi01

  • Forum Obsessive
  • ***
  • Posts: 205
  • Respect: +8
Re: Standard Deviation
« Reply #1 on: October 29, 2018, 05:53:33 pm »
+1
Yo Zoe!

This has been addressed multiple times in the 3/4 question thread recently

Standard deviation refers to how spread the data is around the central value (Mean). Therefore if you have a large standard deviation this means you have a large range of variability in the results. This reduces reliability because it decreases likelihood that if experiment were to be repeated again similar results would be produced.
Furthermore, a large sd could suggest there is not a relationship between the two variables
I think KiNSKi01 adressed the first part of your question really well, but I just wanted to add something to your second question. If you have a look at the key science skills there is a dot point that says something like 'explain the merit of replicating procedures and the effect of sample sizes in obtaining reliable data' So I think while sample size is related to external validitity the SD wants us to focus on large samples in relation to reliability. Furthermore, the two are related, as in something cannot be externally valid without being reliable. And controlling confounding variables (which is what a large sample size does - or attempts to do) is an important part of both validity and reliability.

taken from from the 3/4 question thread

MOD EDIT: please try not to double post. You can amend your posts to add more information by clicking the button that says modify
« Last Edit: October 29, 2018, 06:03:42 pm by vox nihili »
'18:
Physics
Psychology

'19:
English Language
Chemistry
Further Maths
Maths Methods