## Experimental Design B/C

Anomaly
Exalted Member
Posts: 593
Joined: February 17th, 2017, 10:46 am
Division: C
State: PA
Location: probably somewhere crying

### Re: Experimental Design B/C

What does the rubric mean in the quantitative data section where it says "Report most relevant data"?

Making a presentation on expd for middle schoolers and I come across that and have no idea what it means.
That's the condensed data table with important data such as averages.
ah alright that makes sense thanks
Orefield MS SO 2015-2018, Parkland HS SO 2019-2020
Medal/Ribbon Count
Invitational: 25
Regional: 16
State: 7
y o i n k s

Jacobi
Exalted Member
Posts: 137
Joined: September 4th, 2018, 7:47 am
State: -

### Re: Experimental Design B/C

What does the rubric mean in the quantitative data section where it says "Report most relevant data"?

Making a presentation on expd for middle schoolers and I come across that and have no idea what it means.
I think it means that you should report data that is relevant to the hypothesis.

For example, if you're doing an XP with ball drop height v. drop time, you wouldn't report bounce height.

pal_420
Member
Posts: 1
Joined: February 24th, 2017, 3:24 pm
State: -

### Re: Experimental Design B/C

What do the rules mean by age-appropriate statistic? Does this mean we can't have more complex stuff like ANOVA?

Jacobi
Exalted Member
Posts: 137
Joined: September 4th, 2018, 7:47 am
State: -

### Re: Experimental Design B/C

What do the rules mean by age-appropriate statistic? Does this mean we can't have more complex stuff like ANOVA?
I think it means that you shouldn't be doing mode or range for Division C.
Who knows? You might want Pearson Correlation.

Also, how would ANOVA even apply to XPD data?

nicholasmaurer
Coach
Posts: 376
Joined: May 19th, 2017, 10:55 am
State: OH
Location: Solon, OH

### Re: Experimental Design B/C

What do the rules mean by age-appropriate statistic? Does this mean we can't have more complex stuff like ANOVA?
I think it means that you shouldn't be doing mode or range for Division C.
Who knows? You might want Pearson Correlation.

Also, how would ANOVA even apply to XPD data?
Many proctors will be unfamiliar with ANOVA, so I would not recommend using it for your statistical analysis. It is certainly not necessary to do so.

At the Division C level, I would generally recommend you choose either Pearson's correlation coefficient (r and r^2) or Student's t-test. These will cover almost every scenario and are easy to calculate using a graphing calculator. It is important that you understand and interpret the meaning of the results, rather than simply reporting them.

Mean and standard deviation will almost always be the best measures of center and spread to report. Occasionally, median might be preferred, but you're unlikely to have a sufficient sample size to know if your data is skewed.
Assistant Coach and Alumnus ('14) - Solon High School Science Olympiad
Tournament Director - Northeast Ohio Regional Tournament
Tournament Director - Solon High School Science Olympiad Invitational

Opinions expressed on this site are not official; the only place for official rules changes and FAQs is soinc.org.

dxu46
Exalted Member
Posts: 798
Joined: April 11th, 2017, 6:55 pm
Division: C
State: MO

### Re: Experimental Design B/C

What do the rules mean by age-appropriate statistic? Does this mean we can't have more complex stuff like ANOVA?
I think it means that you shouldn't be doing mode or range for Division C.
Who knows? You might want Pearson Correlation.

Also, how would ANOVA even apply to XPD data?
Many proctors will be unfamiliar with ANOVA, so I would not recommend using it for your statistical analysis. It is certainly not necessary to do so.

At the Division C level, I would generally recommend you choose either Pearson's correlation coefficient (r and r^2) or Student's t-test. These will cover almost every scenario and are easy to calculate using a graphing calculator. It is important that you understand and interpret the meaning of the results, rather than simply reporting them.

Mean and standard deviation will almost always be the best measures of center and spread to report. Occasionally, median might be preferred, but you're unlikely to have a sufficient sample size to know if your data is skewed.
What about for division B? A lot of the statistics would be much more simpler with an algorithm on a graphing calculator, but that's taken away.

UTF-8 U+6211 U+662F
Exalted Member
Posts: 1518
Joined: January 18th, 2015, 7:42 am
Division: C
State: PA

### Re: Experimental Design B/C

I think it means that you shouldn't be doing mode or range for Division C.
Who knows? You might want Pearson Correlation.

Also, how would ANOVA even apply to XPD data?
Many proctors will be unfamiliar with ANOVA, so I would not recommend using it for your statistical analysis. It is certainly not necessary to do so.

At the Division C level, I would generally recommend you choose either Pearson's correlation coefficient (r and r^2) or Student's t-test. These will cover almost every scenario and are easy to calculate using a graphing calculator. It is important that you understand and interpret the meaning of the results, rather than simply reporting them.

Mean and standard deviation will almost always be the best measures of center and spread to report. Occasionally, median might be preferred, but you're unlikely to have a sufficient sample size to know if your data is skewed.
What about for division B? A lot of the statistics would be much more simpler with an algorithm on a graphing calculator, but that's taken away.
I would say mean and standard deviation are fairly easy to calculate by hand, but you could also use the mean absolute deviation which takes the absolute values instead of squaring and then square rooting (even easier by hand!).

Jacobi
Exalted Member
Posts: 137
Joined: September 4th, 2018, 7:47 am
State: -

### Re: Experimental Design B/C

I think it means that you shouldn't be doing mode or range for Division C.
Who knows? You might want Pearson Correlation.

Also, how would ANOVA even apply to XPD data?
Many proctors will be unfamiliar with ANOVA, so I would not recommend using it for your statistical analysis. It is certainly not necessary to do so.

At the Division C level, I would generally recommend you choose either Pearson's correlation coefficient (r and r^2) or Student's t-test. These will cover almost every scenario and are easy to calculate using a graphing calculator. It is important that you understand and interpret the meaning of the results, rather than simply reporting them.

Mean and standard deviation will almost always be the best measures of center and spread to report. Occasionally, median might be preferred, but you're unlikely to have a sufficient sample size to know if your data is skewed.
What about for division B? A lot of the statistics would be much more simpler with an algorithm on a graphing calculator, but that's taken away.
Given that they say "age-appropriate", they would expect mean, median, range, mode???? , maybe Standard Deviation. Those formulae are easy to memorize and execute on a 4-function or non-programmable scientific calculator.

dxu46
Exalted Member
Posts: 798
Joined: April 11th, 2017, 6:55 pm
Division: C
State: MO

### Re: Experimental Design B/C

Many proctors will be unfamiliar with ANOVA, so I would not recommend using it for your statistical analysis. It is certainly not necessary to do so.

At the Division C level, I would generally recommend you choose either Pearson's correlation coefficient (r and r^2) or Student's t-test. These will cover almost every scenario and are easy to calculate using a graphing calculator. It is important that you understand and interpret the meaning of the results, rather than simply reporting them.

Mean and standard deviation will almost always be the best measures of center and spread to report. Occasionally, median might be preferred, but you're unlikely to have a sufficient sample size to know if your data is skewed.
What about for division B? A lot of the statistics would be much more simpler with an algorithm on a graphing calculator, but that's taken away.
I would say mean and standard deviation are fairly easy to calculate by hand, but you could also use the mean absolute deviation which takes the absolute values instead of squaring and then square rooting (even easier by hand!).
Alright, thanks.

Just wondering, are there any ES's out there on the forums? I'm curious on how certain sections are judged (i.e. Analysis, Qualitative, Statistics)

UTF-8 U+6211 U+662F
Exalted Member
Posts: 1518
Joined: January 18th, 2015, 7:42 am
Division: C
State: PA

### Re: Experimental Design B/C

What about for division B? A lot of the statistics would be much more simpler with an algorithm on a graphing calculator, but that's taken away.
I would say mean and standard deviation are fairly easy to calculate by hand, but you could also use the mean absolute deviation which takes the absolute values instead of squaring and then square rooting (even easier by hand!).
Alright, thanks.
(Also see Jacobi's post, although I'm skeptical of range being a good indicator of anything, and you probably won't have a mode)