The Student Room Group

Maths Statistics help Urgent!!!

Hi everyone, I have been given a math assignment to do and have no idea where to start. Firstly, could somebody tell me the difference between Data analysis and hypothesis testing. Secondly, is it possible to compare discrete and continuous data? And finally, what type of summary statistics can I do on data. For example if I have average crime rate and average wage, what can I do as summary statistics. Thanks for all help :smile:
(edited 3 years ago)
You can find this all out for yourself on Google. Getting randomers on TSR to answer won't help you learn. Asking questions on TSR is good for a specific area you're stuck on where some guidance / commentary will help.
Original post by KaptainCliff
You can find this all out for yourself on Google. Getting randomers on TSR to answer won't help you learn. Asking questions on TSR is good for a specific area you're stuck on where some guidance / commentary will help.

I quite clearly couldn't find a lot of it on google which is why I asked on this forum. Also if you actually read the post you will realise that I didn't ask any specific questions, but rather general ones.
Reply 3
Original post by BioChemMathGman
I quite clearly couldn't find a lot of it on google which is why I asked on this forum. Also if you want to get off your high horse and actually read the post you will realise that I didn't ask any specific questions, but rather general ones.

Well, if you ask general questions we can only give general answers. So here we go...:

> the difference between Data analysis and hypothesis testing?

Data analysis is any method that analyses data. Hypothesis testing is a more specific technique in which a controlled setting such as an experiment is used to generate data about a pre-conceived hypothesis which is then tested for using a statistical test, which is specifically chosen to suit the purpose and nature of the experiment.

>is it possible to compare discrete and continuous data?

Sure, you can compare anything you want, such as apples with oranges. Here I am not sure what you even mean with the word "compare" in this context so I cannot give any further answer than this.

> what type of summary statistics can I do on data?

Which summary statistics do you know of yourself? Those are all used to summarize data, so the answer would be all of them. That does not mean they would necessarily make any sense.

Hope that helps!
Original post by just_m3
Well, if you ask general questions we can only give general answers. So here we go...:

> the difference between Data analysis and hypothesis testing?

Data analysis is any method that analyses data. Hypothesis testing is a more specific technique in which a controlled setting such as an experiment is used to generate data about a pre-conceived hypothesis which is then tested for using a statistical test, which is specifically chosen to suit the purpose and nature of the experiment.

>is it possible to compare discrete and continuous data?

Sure, you can compare anything you want, such as apples with oranges. Here I am not sure what you even mean with the word "compare" in this context so I cannot give any further answer than this.

> what type of summary statistics can I do on data?

Which summary statistics do you know of yourself? Those are all used to summarize data, so the answer would be all of them. That does not mean they would necessarily make any sense.

Hope that helps!


Hey man,

I really appreciate this, it was just what i needed. For the second point, I wanted to know whether I could plot a scatter graph with these 'different' types of data. And for summary statistics, I have been told to briefly describe the data and a google search told me to use summary statistics. I am struggling to understand how would you briefly compare data such as education rate per 1000 and average wage. Am I right in saying I can't use graphs such as box plots? Other than the statistical tests such as chi-squared, t-tests, etc, what can be used to summarise. Sorry if I am taking the biscuit I know I am asking a lot of questions, I just can't get around how to compare to seemingly completely unrelated bits of data
Reply 5
H

Original post by BioChemMathGman
Hey man,

I really appreciate this, it was just what i needed. For the second point, I wanted to know whether I could plot a scatter graph with these 'different' types of data. And for summary statistics, I have been told to briefly describe the data and a google search told me to use summary statistics. I am struggling to understand how would you briefly compare data such as education rate per 1000 and average wage. Am I right in saying I can't use graphs such as box plots? Other than the statistical tests such as chi-squared, t-tests, etc, what can be used to summarise. Sorry if I am taking the biscuit I know I am asking a lot of questions, I just can't get around how to compare to seemingly completely unrelated bits of data

Hi, from your writing it is clear to me that you don't really understand the difference between different data types such as continuous, discrete, categorical, or which types of charts are best to show each type (such as line chart, box plot, histogram). That is fine, but obvs. I won't go and do your research for you, and I don't know how much time you want to spend on this. I think you'll need to Google a bit until you find a site that explains the basic ideas of the most common chart types and then you'll figure out which one suits your situation best.

Your specific example: education rate per 1000 VS average wage. You use the term "compare" again. I think you may be trying to say you are interested in their relationship. If that's correct, why not try a line chart with one variable on the x-axis and another on the y-axis?

Statistical tests are not use to summarise. They are used to test hypotheses.

You could have 10,000 boys and 10,000 girls and you measure their SAT scores in English and Maths. Then you have 2x2x10.000 data points. Probably too much to plot. So you summarize them into 4 numbers: average English and Maths scores for the boys and girls. Those 4 numbers are the summaries.

" just can't get around how to compare to seemingly completely unrelated bits of data"

Data are not related or unrelated to each other, they are just the numbers representing phenomena. It is the phenomena which may or may not be related, but even if they are related, a statistical test may not be able to detect that (or a statistical test may claim that they are related even though they are not). For example, check out these highly correlated but totally independent data: https://www.tylervigen.com/spurious-correlations
Original post by BioChemMathGman
Hi everyone, I have been given a math assignment to do and have no idea where to start. Firstly, could somebody tell me the difference between Data analysis and hypothesis testing. Secondly, is it possible to compare discrete and continuous data? And finally, what type of summary statistics can I do on data. For example if I have average crime rate and average wage, what can I do as summary statistics. Thanks for all help :smile:

Surely this must link to what you've been learning? What year are you in?
(edited 3 years ago)
Original post by just_m3
H


Hi, from your writing it is clear to me that you don't really understand the difference between different data types such as continuous, discrete, categorical, or which types of charts are best to show each type (such as line chart, box plot, histogram). That is fine, but obvs. I won't go and do your research for you, and I don't know how much time you want to spend on this. I think you'll need to Google a bit until you find a site that explains the basic ideas of the most common chart types and then you'll figure out which one suits your situation best.

Your specific example: education rate per 1000 VS average wage. You use the term "compare" again. I think you may be trying to say you are interested in their relationship. If that's correct, why not try a line chart with one variable on the x-axis and another on the y-axis?

Statistical tests are not use to summarise. They are used to test hypotheses.

You could have 10,000 boys and 10,000 girls and you measure their SAT scores in English and Maths. Then you have 2x2x10.000 data points. Probably too much to plot. So you summarize them into 4 numbers: average English and Maths scores for the boys and girls. Those 4 numbers are the summaries.

" just can't get around how to compare to seemingly completely unrelated bits of data"

Data are not related or unrelated to each other, they are just the numbers representing phenomena. It is the phenomena which may or may not be related, but even if they are related, a statistical test may not be able to detect that (or a statistical test may claim that they are related even though they are not). For example, check out these highly correlated but totally independent data: https://www.tylervigen.com/spurious-correlations

Ok I think I have been approaching the problem wrong. I think it would be better to analyse my data separately and hence use the graphing methods specific to each type of data as I have been told to 'briefly summarise my data' before using the statistical tests to show whether there is an agreement to my hypothesis. Please correct me if that sounds wrong but thanks again for your helpful reply
Original post by Muttley79
Surely this must link to what you've been learning? What year are you in?

I'm in year 13. The idea I was struggling to grasp is how to 'summarise' data as i have been told to without using statistical testing.
Original post by BioChemMathGman
I'm in year 13. The idea I was struggling to grasp is how to 'summarise' data as i have been told to without using statistical testing.

Summarise at A level usually means find means, sd etc for the data using an appropriate method. Is this to do with the large data set?
Original post by Muttley79
Summarise at A level usually means find means, sd etc for the data using an appropriate method. Is this to do with the large data set?

I have 47 values and 2 variables so I'd assume so. Are there any higher level summary methods I could research and use as I have been told to go beyond my syllabus
Which board is this?
Original post by BioChemMathGman
I have 47 values and 2 variables so I'd assume so. Are there any higher level summary methods I could research and use as I have been told to go beyond my syllabus
Original post by Muttley79
Which board is this?

I do Pre-u. This is a separate project though. Is my sample size large enough to be normally distributed by the central limit theorem?
Original post by BioChemMathGman
I do Pre-u. This is a separate project though. Is my sample size large enough to be normally distributed by the central limit theorem?

I'm not familiar with that spec so check what is in the FMaths to see what else you could do - it all sound very vague :smile:

If it's not part of the examjust do what you can.

Quick Reply

Latest