You, my dear man, are a lifesaver. I easily spent an hour trying to figure this out on my own and I imagine I would have spent far more time if I didn't find your video.
I am currently completing a scientific project for my dissertation which will be submitted by the end of August this year! I honestly did not understand one-way or two-way ANOVA from the explanations that were given to us and after watching this I did it with ease! A massive thank you - I cannot explain how much help this has been :)
Ay man i love you! I have watched your Chi square tutortial before during my bachelor in environemental sciences. And now im writting a bachelor in biology! Irish accent is also very charming
Sir, please enlighten me about what you said in 5:40 minutes. You said the p-value (5.4E-08) is tiny which means, reject the null hypothesis. I am confused, how can it be tiny compared to the 0.05 alpha value? I will discuss this topic, and your video greatly helps me. Only that, I am confused about this part. Thank you.
I'm enlightened already sir. This might help others who have the same confusion with 5.4E-08. At first, I assumed this was greater than the alpha value 0.05. I asked the help of ChatGPT and here is the explanation: The notation "5.43985E-08" is a representation of a number in scientific notation. It's a way to express very large or very small numbers in a compact form. In scientific notation, a number is expressed as a coefficient multiplied by a power of 10. The "E" stands for "exponent," and it tells you how many places to move the decimal point. So, "5.43985E-08" means 5.43985×10 −8 , which is equivalent to 0.0000000543985. Indeed, this is very tiny compared to the 0.05. Hope this helps.
Hi TechByte, You should not get both together - please recheck your calculations. An F stat greater than F crit means that you reject the Null Hypothesis, and p-value greater than alpha means that you cannot reject the Null Hypothesis. Hope this helps, Dr E.
@@EugeneOLoughlin thanks for the reply Dr. E! Really appreciate it :) I’ve since found the mistake: I had a typo in my alpha value. It was set to 0,5 instead of 0,05 🤦♂️ Still interesting though, that the p-value and F statistic can contradict each other when you have a large alpha! By the way, are there any other follow up tests you can recommend if none of your null hypotheses can be rejected? My p-values are 0,478, 0,435 and 0,078 respectively, so there are no statistically significant effects or interactions at a level of alpha = 0,05. Thank you, and have a nice day!
Hi Eugene. Thank you for all "How to..." videos, are REALLY helping me. One quick question: the TWO WAY ANOVA test can be performed by hand also rather than performed in Excel? Thank you
Hi Alexandra, First - many thanks for your kind comment. The 2 x ANOVA, and all other ANOVA tests, can of course be done by hand. But it is a lot more complicated then the 1 x ANOVA. I don't cover it in my classes - I just use Excel (plus SPSS and R) to perform the two-way ANOVA. Dr E.
I Dr. I ran the Two Way Anova with replication but, I got an output with a = #NUM! and p-value = #NUM!. I also imported the data into spss. The result came out without interaction. What can I do in this case Sir?
Sir I have a dependent variable i.e accident count in form of numbers and multiple independent variables i.e different land use in form of areas , please suggest which statistical analysis can be used. I have these data for 7 different areas .
Hi Eugene, is it the case that what Excel understands by "with replication" is what is traditionally known as "between subjects"? In other words, is the analysis you explain here suitable if the different rows in each category represent different individuals (cases), and the category is a type of individual, e.g. instead of the Super and Best brands with its 4 rows each, we had Men and Women categories, each with 4 subjects? many thanks!
My understanding is that 'between subjects' means each person gets only 1 treatment while 'within subjects' can be where the same person gets more than 1 treatment. I don't have my statistics book with me but the tricky part is whether the number of subjects for each treatment has to be the same. Here, it seems to be yes.
You have not specified that which 2 are different It is done by finding Critical difference and Mod(T1-T2) T1 and T2 are averages of brands Please make a separate video for this problem.
Hi Mayar, Unfortunately Excel 2013 does allow for uneven data - it will not work. Don't be tempted to insert zeros or other data, as this will mess with your results. You may need to work out by hand/manually, or use another tool such as SPSS. Dr E.
Hi Joshua, This error is usually due to one of two things. First - if you have any non-numeric data in your data table, like a character or date, this will not work with any ANOVA tool. So the obvious first check is to remove any non-numeric data. But the second, and most common, reason is that you do not have the same number of values in each cell. In my example in the video there are four values in each cell: eg, four Cold values for the Super brand, four Warm values for the Super brand, and so on. In Excel you MUSThave the same number of values in each row - in my video at 2:35 I enter the number 4 in the "Rows per sample" box. So if you have three values in one cell, and four in the rest - Excel will not work. This is a quirk of Excel - other tools such as SPSS can handle this. Hope this helps, Dr E.
@@EugeneOLoughlin I checked so many times and it really is not working. The only thing I can notice on the sheet Im working on is that every number is in 2 decimal places (e.g. 14.20). The headings are non numeric but it worked for you when youused the ANOVA function on the table you have in the video. I also rechecked and it has equal number of rows per sample. I don't really know now whats wrong with excel
Hi Joshua, Two decimal places should not matter. Double check that all your data are formatted as numeric, and that you have laid it out on the spread sheet in the way my data are at the beginning of the video. Make sure that you do not have any blank cells. Otherwise I cannot tell why this is not working for you. Dr E.
