One-Way ANOVA with LSD (Least Significant Difference) Post Hoc Test in Excel

This video demonstrates how to conduct an ANOVA with a Fisher’s Least Significant Difference (LSD) post hoc test in Microsoft Excel. A comparison of the LSD in Excel is made to the SPSS output.

[lordsiomai]

You have no idea how much your videos help students like me with their theses

[jewelroman]

Thank you so much, Our professor didn't want us to see this. He wants us to solve manually but the time he gave us to solve it manually is just unrealistic that we always got low grades. :<

[lawrence]

Your videos are always helpful, thank you Dr. Grande!

[omniaarief]

OMG thank u very much, U have no idea how much I needed this video, I'm trying to calculate the LSD value by R and sadly didn't get a working equation and this video is a life-saving. May Allah bless u :)

[genarahmmahato]

The ANOVA table used all three tretments for calculating Within Groups MS. However, LSD formula used only the number of observation of CBT and Group, not Treatment as Usual. Why would we use only the number of observation of first two treatment for calculation of LSD, and not use number of observation of other treatments? Consider, if we had total of 10 treatments, why would we use only the number of observation of first two treatments ( 1/N1 and 1/N2), and omit the number of observation of other 8 treatments, for calculation of LSD. Would you please help me to find out the answer of this question.

[khandkerhaque]

Great demonstration in short video. Thanks very much.

[Ethan_Stonecipher]

Thank you for the video, Dr. Grande. This helped me tremendously!

[Zenith-hrut]

So there is a formula for LSD where t value (critical) is divided by 2 and in you LSD formula it isn't.. so what's the difference?

[Solochi]

For the critical value of t, I typed "=t.inv", but it is not showing up😢

[OlwethuSindesi]

Good day

This has been helpful.

However, I am not from a statistic background I am an agronomist and normally just outsource.

I would like to find out how I can then use this significant value in drawing my graphs.

For instance, the mean values are made clear by the ANOVA and I can easily pick them and draw my graph showing the various means from my treatments. Now I would like to put asterisks on graphs to show which one did better than the other and which did poorly.

Can I then just look at the largest mean then put an "a" asterisk on top of the bar, then subtract the LSD value and find which treatment (I have 4 treatments) mean is closest then denote it with a "b" and so forth until my treatments are all covered?

Please I have minimal background in stats.

[raywinovah]

What if the p-value is greater than 0.05? Is there other methods to use when the test is not statistically significant at 0.05?

[asagrimmr]

What does it mean to have a "statistical significance"? Is that rejecting that the two means equal or not?

[rmwxcne]

can anyone tell what shall i write instead of NA in excel, since NA is not accepted by ANova test selection, and if i put zero"0" then it may cause the change of overall test in averages.

[marianadeiak]

Hello Dr. Todd Grande,
Thanks for the explanation!
How much is it relevant to compare the data in scientific work with huge differences in parameters? Because then abs = 250!!
Is it possible that LSD = 8 and another one LSD = 50. For me it seams to much 50




This is my data to understand easier

1000-grain weight, g Yield, t/ha
227, 8 5, 3

285, 3 6, 1