Hello everyone

#1
My name is Iwan Bonnén

I'm from Copenhagen, Denmark.
I have a Master's in Mathematics and am currently working on my project www.medicalstat.org which is an online app within medical statistics, biostatistics and epidemiology.

I'm primarily looking for people who can evaluate my project and give me feedback on the app: what is good/bad, what features are missing ...
I'm also looking for people who would like to support the project, both financially but also by testing the features and give ideas to design, new functionalities etc.

Best regards
Iwan
 
#3
Hi there.

Thanks a lot :)

Well, it's not actually a "software" in the sense that it cannot be downloaded or installed.
It's an online application to do statistical calculations on site, without download or installation.
That way I try to keep it easily accessible and user friendly.

It's written in the standard web languages: HTML, JavaScript, CSS.
 
#5
Isn't there enough free software already?
That depends on what you mean with "enough" in this context.
Yes there is A LOT of free software and websites, I know that. But within this particular field (medical statistics) there are not overwhelmingly many.
And some of the existing ones lack a few features, which I hope to include on my site.
 
#7
Well, by trying it out. By doing some calculations with it. Perform a z-test, t-test, f-test, upload data from a text file, plot some diagrams ...
And see if it meets their expectations. Is it user friendly enough? What features are your missing ... etc.
 

hlsmith

Less is more. Stay pure. Stay poor.
#8
The site seemed fine. The following are a couple of comments. Aesthetically - the page looks basic and not as though it was made recently. I know this is trivial - but a general comment about appearances. The font seems small, so I increase the viewing size and it distorts the layout, etc. Some issues there for me related to readability and scaling size.

It is not apparent that the site serves to plug-n-play when on the home page. Perhaps that should be better conveyed.

I didn't look that hard but you may need to define more things (e.g., acronyms) the kind of people that would look for a website to do their analyses are likely not gonna know what you mean by SE. so term definitions, etc. may be needed. There may not be enough background info on topics. The rule of five, I am guessing is in regards to using X^2 with small expectancies - by your viewers likely need more info on this. Sidenote I think that rule is not actually based on any data - and the general recommendation could be to always use Fisher's given the process can execute in reasonable time.

Good luck!
 
#9
Thanks for the feedback, hlsmith.

Yes, I know about the design. I'm planning on changing it in the near future. I'll look into the font size. Design and layout are not my strong sides ^^

I didn't look that hard but you may need to define more things (e.g., acronyms) the kind of people that would look for a website to do their analyses are likely not gonna know what you mean by SE. so term definitions, etc. may be needed. There may not be enough background info on topics.
I tried to solve this exact issue by including the section "formulas", where I not only explain the abbreviations, but also try to explain the concepts and how they're calculated. Sometimes it would take too much space to write "standard error" in full everywhere. But actually I'm planning on including a few notes on medical statistics including some examples.

Regarding the abbreviations SE, SD, ... etc. My target audience is actually not people who are completely "green" to statistics. I expect that people who could find use of my app have at least had an introductory course in statistics, medical statistics or similar, where terms like "SE" are used almost per default. My approach was that someone who could find use of my site would also be someone who knew what SE stands for, but I might reconsider.
 

Dason

Ambassador to the humans
#10
It's never a bad idea to write out the full term the first time it's introduced and provide the abbreviation so you can use that afterwards. You only need to write the term once but it eliminates confusion. Even for audiences familiar with the topic sometimes there are abbreviations that have multiple meanings and it's always beneficial to be clear which you mean.
 

hlsmith

Less is more. Stay pure. Stay poor.
#11
ML = machine learning or maximum likelihood; CI = confidence interval, causal inference, maybe credible inference, etc.

I would argue that anyone that knows statistical terms would be more likely to just run these procedures on their own. This way they have reproducible code that can be referenced again in the future if needed. And that you will likely get a person like a nursing student trying to run a test for the first and only time. Where that person will be ignorant to the appropriate application and assumptions.

Just my 2 cents.