Really stuck with a hypothesis

Ave Sodalis · 22-01-2015 6:06pm #1

Hi guys,
Really need some help with this. For my FYP, I'm expected to come up with a hypothesis. This ordinarily would be completely fine if I had any control over the experiment. However, my supervisor keeps changing the experiment any time we encounter any sort of problem which is doing my head in. Anyway, as I haven't been allowed any input into my experiment, I've very little understanding of it. We've finally agreed on a title a few weeks ago. However, when I sat down to research a literature review (one of many at this stage), I realised that I've no hypothesis. I've no idea what I'm actually supposed to be trying to find out. The title is "Comparison of Different Aspects of Physical Fitness in a Student Population in Relation to Equestrian Specific Tasks." and for the experiment, I'm running several fitness testing along with getting the subjects to fill out an extensive profile which goes into everything from their sporting background to what they ate for breakfast that morning.

I have no idea what I can find out from it though. I'm Equine Science so it has to be equine based. Majority of the participants will be from an equine background, with some participants who aren't. One of the original ideas was to catagorise each participant into their training (Equine Sci, BHS exams, RACE, non equine student) and compare fitness levels. However, there were problems with this (i.e. the rest of my lecturers said I should just use Equine Sci students and compare them as equine athletes against non equine athletes but for some reason, my supervisor insisted that I use all groups and told me not to do a comparison if it involved leaving them out).

So now I'm at a loss. I've had no control over anything and not allowed to make a decision about it and now she wants me to find a hypothesis that she agrees with but is ignoring my emails asking for help. So I'm turning to you before I go insane. Any idea what I could do with a bunch of fitness results?

Black Swan · 23-01-2015 6:37am

sup_dude wrote: »

I realised that I've no hypothesis. I've no idea what I'm actually supposed to be trying to find out. The title is "Comparison of Different Aspects of Physical Fitness in a Student Population in Relation to Equestrian Specific Tasks." and for the experiment, I'm running several fitness testing along with getting the subjects to fill out an extensive profile which goes into everything from their sporting background to what they ate for breakfast that morning.

Research Hypothesis:

There is a relationship between different aspects of physical fitness and equestrian specific tasks.

***

But we do not test the Research Hypothesis, rather by convention we test the

Null Hypothesis:

There is no relationship between different aspects of physical fitness and equestrian specific tasks.

Ave Sodalis · 23-01-2015 11:40am

See the equine specific tasks are the fitness tests. For example, number of sit ups in a certain amount of time which tests core strength, needed for horse riding. Or an ankle flexibility test which is being used because equestrians need ankle flexibility.

innad · 23-01-2015 11:57am

Students from an equine background perform better than those not from an equine background on equine specific fitness tests.

As Black Swan says, you would test the null hypothesis:
there is no difference between the various student groups on how well they perform on the equine specific fitness tests.

Ave Sodalis · 23-01-2015 12:05pm

innad wrote: »

Students from an equine background perform better than those not from an equine background on equine specific fitness tests.

As Black Swan says, you would test the null hypothesis:
there is no difference between the various student groups on how well they perform on the equine specific fitness tests.

That was an earlier idea but my supervisor said no. She wanted to use all the equine groups and didnt want to explain to the other lecturers (or me) why I can't just use one equine group so comparisons are out.

It was also an idea to find out if there was a relationship between fitness levels and performance on the racehorse simulator, but that was discarded as my supervisor came under a lot of questioning when I presented it to the other lecturers.

innad · 23-01-2015 12:29pm

Is there any way you can consult one of the other lecturers if your supervisor is ignoring you?

At the end of the day it's your project, your supervisor can advise you but you're really under no obligation to follow their advice.

challengemaster · 24-01-2015 12:54am

sup_dude wrote: »

However, my supervisor keeps changing the experiment any time we encounter any sort of problem which is doing my head in.

Welcome to Science

Anyway, as I haven't been allowed any input into my experiment, I've very little understanding of it. We've finally agreed on a title a few weeks ago. However, when I sat down to research a literature review (one of many at this stage), I realised that I've no hypothesis. I've no idea what I'm actually supposed to be trying to find out.

I'd say the fact you have very little understanding of the project yet has a lot to do with that. A hypothesis is a proposed explanation for something. Ideally you shouldn't just pluck them out of thin air, but instead base a hypothesis on already existing research and findings.

It might be easier to give an (imaginary) example - In your literature review you find a load of papers relating to diet and corresponding ankle flexibility. Imagine these are already published findings. You could then form a hypothesis that diet is a contributory factor in equine specific task performance.

There's one important thing I feel the need to point out, which may be holding you back based on this:

I have no idea what I can find out from it though.

Your hypothesis does not have to be correct. There is a severe bias in science so that the majority of papers published (particularly in high impact factor journals) are all positive results. Fact is a result is a result, be it positive or negative. Younger/inexperienced researchers tend to think of the outcome and almost try to plan the experiment to get their intended outcome. You want to know what you can find out from it? Do the experiment.

One of the original ideas was to catagorise each participant into their training (Equine Sci, BHS exams, RACE, non equine student) and compare fitness levels. However, there were problems with this (i.e. the rest of my lecturers said I should just use Equine Sci students and compare them as equine athletes against non equine athletes but for some reason, my supervisor insisted that I use all groups and told me not to do a comparison if it involved leaving them out).

sup_dude wrote: »

That was an earlier idea but my supervisor said no. She wanted to use all the equine groups and didnt want to explain to the other lecturers (or me) why I can't just use one equine group so comparisons are out.

Off the top of my head, 2 reasons I can think of would be population bias and statistical significance (you're reducing your n number significantly by cutting out other groups).

A third reason would be it's not the aim of your project.

Comparison of Different Aspects of Physical Fitness in a Student Population in Relation to Equestrian Specific Tasks

You're not comparing different student populations. You're comparing physical fitness levels in one student population.

It sounds to me like you may have misunderstood the supervisor on one thing though. I'm reading it that you should:

Carry out the experiment(s) on all of the various populations you mentioned
Compare those results to see what you find out. Just remember what you're actually looking for.

If you collect all the data correctly you can still group results after all is said and done, and see do you find any correlation (eg. The sub-population of Equine Sci students performed better at Equine specific tasks; however this may be due to better understanding of the tasks involved/likelihood of experience with equine specific tasks/etc)

innad wrote: »

At the end of the day it's your project, your supervisor can advise you but you're really under no obligation to follow their advice.

Eh.... You really are. Considering for a FYP your supervisor is one of the people correcting it and grading you as a student. Not following their advice is a sure way to get a bad grade.

innad · 24-01-2015 10:50am

challengemaster wrote: »

Eh.... You really are. Considering for a FYP your supervisor is one of the people correcting it and grading you as a student. Not following their advice is a sure way to get a bad grade.

Obviously in the vast majority of cases this is true. However, it doesn't sound like this supervisor is actually giving any advice. You are allowed to bring your own independent thoughts and ideas, as long as you can back them up with theory and/or evidence.

I do agree that you don't need to know what you're doing to get a good handle on the literature though - a good understanding of the existing research will make it much easier for you to understand what is going on with the experiment, and what the various options might be. You might find potential hypotheses from reading previous studies. Then you could say to your supervisor, "X et al did Y and found Z, but they didn't look at ABC, I think I could use the experiment to look at this, what to do you think?" Then if they say no, ask them to explain the reasons for this, that you think it would really help your learning if you understood more about why different options aren't considered feasible by your supervisor etc.

Ave Sodalis · 24-01-2015 11:55am

challengemaster wrote: »

Welcome to Science

Which would be fine if it was "oh that's not going to work, we'll have to change it". Right now though it's "I want this done for next week" Next week "What did you do that for? Sure that's just a waste of time!"

I'd say the fact you have very little understanding of the project yet has a lot to do with that. A hypothesis is a proposed explanation for something. Ideally you shouldn't just pluck them out of thin air, but instead base a hypothesis on already existing research and findings.

It might be easier to give an (imaginary) example - In your literature review you find a load of papers relating to diet and corresponding ankle flexibility. Imagine these are already published findings. You could then form a hypothesis that diet is a contributory factor in equine specific task performance.

There's one important thing I feel the need to point out, which may be holding you back based on this:

Your hypothesis does not have to be correct. There is a severe bias in science so that the majority of papers published (particularly in high impact factor journals) are all positive results. Fact is a result is a result, be it positive or negative. Younger/inexperienced researchers tend to think of the outcome and almost try to plan the experiment to get their intended outcome. You want to know what you can find out from it? Do the experiment.

That's the problem.The experiment is just a load of fitness tests. There's no direction or basis for any of it. From what I can see, there isn't anything I can use any result of it for.

Off the top of my head, 2 reasons I can think of would be population bias and statistical significance (you're reducing your n number significantly by cutting out other groups).

A third reason would be it's not the aim of your project.

Just using one equine group shouldn't have too much of an effect on statistical significance as the group proposed as I can increase the numbers of that group easily. Plus, another argument for using just Equine Science is many have experience of or even gone through the other two equine courses.

I don't know what the aim of my project is.

You're not comparing different student populations. You're comparing physical fitness levels in one student population.

It sounds to me like you may have misunderstood the supervisor on one thing though. I'm reading it that you should:

Carry out the experiment(s) on all of the various populations you mentioned
Compare those results to see what you find out. Just remember what you're actually looking for.

This is my problem. I don't know what it is I'm looking for.

If you collect all the data correctly you can still group results after all is said and done, and see do you find any correlation (eg. The sub-population of Equine Sci students performed better at Equine specific tasks; however this may be due to better understanding of the tasks involved/likelihood of experience with equine specific tasks/etc)

Again, that was a previous experiment idea. To compare the results of the equine groups but again, that was met with a no.

Eh.... You really are. Considering for a FYP your supervisor is one of the people correcting it and grading you as a student. Not following their advice is a sure way to get a bad grade.

Which is really bad, considering this is the first year my supervisor has done this (only second year as a lecturer). She pretends she knows exactly what she's talking about, gets angry if I correct her on anything, gets corrected by another lecturer and then act as if it was my fault she was wrong the whole time. I really feel like just giving up on it, because we're going around in circles more than getting anything done.

innad wrote: »

Is there any way you can consult one of the other lecturers if your supervisor is ignoring you?

I've tried. I've gone to both the course director and faculty head. Neither could do much as they don't understand what my supervisor is trying to do either.

Black Swan · 24-01-2015 10:34pm

sup_dude wrote: »

Which would be fine if it was "oh that's not going to work, we'll have to change it". Right now though it's "I want this done for next week" Next week "What did you do that for? Sure that's just a waste of time!"

I don't know what the aim of my project is.

This is my problem. I don't know what it is I'm looking for.

I've tried. I've gone to both the course director and faculty head. Neither could do much as they don't understand what my supervisor is trying to do either.

Have you consulted the dissertation abstracts for similar studies? Perhaps you will find one that relates to your titled area of interest? If so, often in dissertation final chapters they have recommendations for future research, which you could adopt and then use as a compelling reason to followup with said research recommendations. Of course, you may have to narrow your scope a bit.

dimsumss · 24-01-2015 10:35pm

hiv in horses is a good one

Ave Sodalis · 24-01-2015 10:42pm

Black Swan wrote: »

Have you consulted the dissertation abstracts for similar studies? Perhaps you will find one that relates to your titled area of interest? If so, often in dissertation final chapters they have recommendations for future research, which you could adopt and then use as a compelling reason to followup with said research recommendations. Of course, you may have to narrow your scope a bit.

I've hoped to try that. No-one on this course has done anything related so far but I'm still searching. I honestly wouldn't mind narrowing my scope, especially concerning the profile for each participant which is three pages long. It's my supervisor I'd have to convince.

dimsumss wrote: »

hiv in horses is a good one

Unfortunately, it's been done to death and completely irrelevant to the testing I'm doing.

Black Swan · 24-01-2015 10:52pm

sup_dude wrote: »

I've hoped to try that. No-one on this course has done anything related so far but I'm still searching. I honestly wouldn't mind narrowing my scope, especially concerning the profile for each participant which is three pages long. It's my supervisor I'd have to convince.

I am not familiar with your discipline's peer-reviewed journals, but sometimes in the conclusions of articles (that may be related to your topic) they too have recommendations for future study. Adopt one of these recommendations and use it to convince your lecturer.

challengemaster · 24-01-2015 11:57pm

sup_dude wrote: »

That's the problem.The experiment is just a load of fitness tests. There's no direction or basis for any of it. From what I can see, there isn't anything I can use any result of it for.

Any experiment is just a load of ___ tests until you give it context. If you take any results section of any research paper and ignore the intro/methods/discussion, you'll have very little idea what they're trying to deduce.

I don't know anything about Equine Science. I doubt most others here do either - so you'll have to help us a bit for us to help you.

What are you defining as Equestrian specific tasks? You've already mentioned ankle flexibility and core strength, so what else? More importantly - what is this actually a measure of? Why is it important in horse riding?

Once you've defined these tasks, your experiment is based around establishing a link between physical fitness and these specific parameters.

So then you have specific tests to measure the equestrian specific tasks you've defined, and a second set of general physical fitness tests and indicators. I don't know what equipment you have access to but you could use: heart rate, body composition(BMI), lung capacity/aerobic capacity (VO2 max), bleep test, etc etc.

There are literally hundreds/thousands of exercise physiology / sports science papers you could look at for your measurements of general physical fitness that are widely used.

Example: https://research.wsulibs.wsu.edu/xmlui/bitstream/handle/2376/4013/B_Elder_010207405.pdf?sequence=1

Plus, another argument for using just Equine Science is many have experience of or even gone through the other two equine courses.

That's exactly what I meant by having a biased population. By exclusively having a pool of participants with experience of what you'll be measuring, you're almost guaranteed to have an artificially positive result which wouldn't be representative of the broader population. Doing that makes the entire research a waste of time.

Your supervisor may want to use the data generated from this project as preliminary data to attempt to get a grant for a larger research project. They have no hope of convincing a funding body by using biased research results.

I don't know what the aim of my project is.

This is my problem. I don't know what it is I'm looking for.

Read the title of your project. That's more or less it.

Your thesis should read like a story that flows.

In your thesis you should have:
1. Background & Introduction (literature review)
1.x Objectives
2. Methods
3. Results
4. Discussion
5. Conclusion
6. References
Appendix

Your background and introduction is where you set the scene. It gives context to the experiment. Why are you going to do this experiment? What effect does performance at equestrian specific task have on ...? These are the types of questions you need to ask yourself when writing your chapters/sub-sections, because that's what anyone else reading it will be wondering.

The last sub-section of your introduction should be your Aims & Objectives. Split this into:

1.x.1. General aim of the study
The overall objective of this study is to evaluate how different aspects of physical fitness in a student population compare to performance ability at equestrian specific tasks.

1.x.2. Specific aims of the study
The specific objectives of this study (how you propose to do this and what parameters you'll measure).

Again, that was a previous experiment idea. To compare the results of the equine groups but again, that was met with a no.

Don't think of it as a separate experiment idea. Think of it as post-hoc analysis that you do once you have all the experiments done. As long as you collect the data (what previous training a person has) at some point (presumably a single question in the survey) you can use it how you like after the fact.

I've tried. I've gone to both the course director and faculty head. Neither could do much as they don't understand what my supervisor is trying to do either.

Have you considered asking the course director to change supervisor? It doesn't sound like you're too invested yet.

Ave Sodalis · 25-01-2015 12:17am

challengemaster wrote: »

What are you defining as Equestrian specific tasks? You've already mentioned ankle flexibility and core strength, so what else? More importantly - what is this actually a measure of? Why is it important in horse riding?

Ankle Flexibility, core strength, hip flexibility, aerobic, dexterity in legs and arms, and muscle endurance. They're needed to sit on a horse, at a guess. There's no evidence that these tasks are needed in a science experiment. There's also a racehorse simulator performance test which is being used as a fitness test. The idea of comparing the other fitness results to the simulator performance results and seeing if there's correlation between fitness and performance was rejected.

Once you've defined these tasks, your experiment is based around establishing a link between physical fitness and these specific parameters.

The parameters are the fitness results.

That's exactly what I meant by having a biased population. By exclusively having a pool of participants with experience of what you'll be measuring, you're almost guaranteed to have an artificially positive result which wouldn't be representative of the broader population. Doing that makes the entire research a waste of time.

The groups of equines I'm using are Equine Sci, BHS and RACE. However, Equinw Sci may have people from BHS and RACE. So surely, by using BHS and RACE, I'd be biasing the results more than if I didn't use them?

Read the title of your project. That's more or less it.

I have, it tells me nothing except what I'm doing for testing.

1.x.1. General aim of the study
The overall objective of this study is to evaluate how different aspects of physical fitness in a student population compare to performance ability at equestrian specific tasks.

That's not the aim though, the equine specific tasks are the fitness tests.

Don't think of it as a separate experiment idea. Think of it as post-hoc analysis that you do once you have all the experiments done. As long as you collect the data (what previous training a person has) at some point (presumably a single question in the survey) you can use it how you like after the fact.

It was a previous idea more than a separate idea. The reason it's a previous idea is because I was told no.

[/QUOTE]Have you considered asking the course director to change supervisor? It doesn't sound like you're too invested yet.[/QUOTE]
I tried but there's nobody else to take me.

challengemaster · 25-01-2015 12:54am

sup_dude wrote: »

The parameters are the fitness results.

The way I'm reading it, you should be comparing two sets of fitness results. One general, the other equine specific.

I have, it tells me nothing except what I'm doing for testing.

That's not the aim though, the equine specific tasks are the fitness tests.

You said in the OP that you came to an agreement on the title. The title should tell you what the entire study is about. Only two things are possible here:

The title is wrong
You have misunderstood the title.

Until you figure out which of those is true and rectify it, you're at an impass which nobody on here can help with.

"Comparison of Different Aspects of Physical Fitness in a Student Population in Relation to Equestrian Specific Tasks."

What that says to me (and clearly others) is you're examining the relationship between physical fitness and ESTs. You've said that's not the case.

"Comparison of Different Aspects of Physical Fitness concerning Equestrian Specific Tasks in a Student Population."

Now I believe that more accurately conveys your understanding of the title to everyone else. However, now the project makes very little sense. Short of comparing each test to another and hoping for correlation the project actually seems to have no discernible aim.

They're needed to sit on a horse, at a guess

Again, you said you came to an agreement on the project title. If you don't know what these tests will actually relate to in terms of horse riding; how will anyone else? More to the point, how did you agree to something you didn't understand?

I tried but there's nobody else to take me.

Approach the course director again. If that doesn't work then go to the dean of science.

Ave Sodalis · 25-01-2015 1:04am

challengemaster wrote: »

The way I'm reading it, you should be comparing two sets of fitness results. One general, the other equine specific.

You said in the OP that you came to an agreement on the title. The title should tell you what the entire study is about. Only two things are possible here:
The title is wrong

You have misunderstood the title.

Until you figure out which of those is true and rectify it, you're at an impass which nobody on here can help with.

"Comparison of Different Aspects of Physical Fitness in a Student Population in Relation to Equestrian Specific Tasks."

What that says to me (and clearly others) is you're examining the relationship between physical fitness and ESTs. You've said that's not the case.

"Comparison of Different Aspects of Physical Fitness concerning Equestrian Specific Tasks in a Student Population."

Now I believe that more accurately conveys your understanding of the title to everyone else. However, now the project makes very little sense. Short of comparing each test to another and hoping for correlation the project actually seems to have no discernible aim.

Again, you said you came to an agreement on the project title. If you don't know what these tests will actually relate to in terms of horse riding; how will anyone else? More to the point, how did you agree to something you didn't understand?

Approach the course director again. If that doesn't work then go to the dean of science.

By agreeing to the title, it was more a case of "this is what your title is going to be, you don't have a say in it, I'm not taking any of your suggestions and that is that". To be honest, I was happy to agree because the rest of them went something like "yeah that's perfect, no bother, great idea" "um, this lecture is asking about this part of the experiment and I wasn't really sure of the answer...." "What? Sure why'd they be asking that? That isn't even part of your project! You need to rethink your title" or in one case, she just literally sent an email saying "you need to rethink your title". There was no explanation, it didn't even sign her name off. Any email I sent to her after that, asking to explain was ignored.

Really stuck with a hypothesis

Comments