How to do a One-Way ANOVA in SPSS (12-6)

[@lisagustafsson2028]

Hi I just wanted to tell you that your 12 minute video is far more valuabel than my SPSS stat.book for about 1200 euros. Your video is covering understanding, coding the data, interpreting and (very important) the APA style. THANKS !!

[@ResearchByDesign]

Wow, that is a really high complement. Thank you so much. Yes, I think that including the example write up in APA style is really important. We need to know how to interpret the stats.

[@rosemaryaanalike]

You video simplified ANOVA!

[@shrutitrivedi4742]

I've been fan of TV shows, movies but this is first time I've been fan of your SPSS videos. You teach so passionately that it really shows. I really have enjoyed and have actually understood so many things than in my real stats class. Thank you so much !

[@Oluwagbeoyemiga]

All thanks to you. I am gonna complete my Msc thesis without any hassle. KZbin has been my greatest teacher. Getting my second degree with much aid from KZbin videos. Yayyy!

[@alikarimi1987]

I do not know how to thank you for making these SPSS videos full of hints and tricks . and the interpretation of data has been of great value . I jotted down the whole video .

[@ResearchByDesign]

Glad you like them! So happy that it has been helpful for you

[@erfanberjisian2492]

I've been always struggling with learning statistics and SPSS but with your valuable videos and simple teaching, I'm learning step by step. thank you for your effort and time putting on these. You are a great professor.

[@telispatroclou7252]

Thank you, speechless! The simplicity in your videos goes in tandem with viewer's IQ showing a positive correlation in learning

[@ResearchByDesign]

Excellent...Glad that it was helpful! I try to make it understandable. Thank you for watching.

[@bidhya9857]

Okay. I paused the video for a minute to say thankyou to you. I watched so many before this but I couldn't understand at all. Seriously THANKYOU so much. Explained everything so simply. Keep it up😊😊

[@ResearchByDesign]

You're very welcome! I'm glad that the videos are so widely viewed and enjoyed. Appreciate your comment

[@Dr.UdaraSenarathne]

Thank you very much! The complete ensemble of test application, interpreting results, and presenting the conclusion was excellently done to prove the viewer a good understanding of Anova.

[@saliknazir4768]

I struggled with this for more than 4 years and you made me understand all of that in just a few videos. I do not know how to thank you. I wish I would have found this channel earlier. But 'All is well that ends well!'. I wish you all the best.

[@chrispaine511]

I WISH YOU WERE MY LECTURER! Thanks for explaining it in a way that is easy to understand. I've watched 15 of your videos today and I feel so much more confident than I did yesterday. I have an SPSS test tomorrow and you saved my bacon! Hazzah!

[@ResearchByDesign]

Great to hear that...glad that the videos were helpful and I wish you the best on your test. Please tell your classmates, too. :o)

[@Volkoj]

WOW, I'm stunned. Couldn't figure it out for weeks. This video helped me so much. You're the best!!

[@ResearchByDesign]

Outstanding! Glad it helped! Thanks for taking the time to comment

[@RUJedi]

I watch your videos as someone who used to tutor stats and enjoys seeing how others coach it. Thanks & kudos on these ANOVA videos. You speak clearly, give interesting examples (nice touch with the wicked witch), and your step-by-step is great. Even better, I appreciate how you translate the SPSS outputs into APA for reporting and demonstrate use of the additional features such as the plots, post-hoc & other tests.

[@ResearchByDesign]

Thanks, Shawn! Really appreciate your kind words. I think it is important to consider the whole research process, not just the tests. You did a great job of explaining how it is supposed to work. Best

[@ResearchByDesign]

Thanks, Shawn...that means a lot, especially coming from someone who tutors stats. Glad that you liked it.

[@sharlenemiller914]

Thank you so much for teaching with such passion I am able to understand. God bless and flourish you. Thank you

[@turgen3416]

YOU ARE THE GREATEST PERSON I'VE EVER SEEN

[@komanekussing6959]

Thank you so much. I have learned in 12 minutes than 3 hour classes i attend at school.

[@ResearchByDesign]

That's great! Best to you in your coursework

[@kavehshafiee]

The best stat course Ive ever had, thank you so much

[@ResearchByDesign]

Thanks for the comment. I read the first version where you accidentally called it a "stat curse" and I thought "how many people have felt that their stats course was a curse?" I love that play on words and I am going to use that with students. Thanks

[@kingmu1]

I want to say thank you for these videos. I am a stats professor and I use your videos in my class!

[@ResearchByDesign]

I love hearing from stats professors! Thank you so much for the comment and my best to you and your students. Thanks for watching the videos.

[@craftychinurse5941]

You are a terrific instructor/professor, teacher! Thank you so much!

[@ResearchByDesign]

You're very welcome!

[@bonfacemacharia7388]

Thank you sir. Your videos have been of tremendous help.

[@bandikelarajesh8554]

Today I have an semester exam on SPSS 😅 and I have found this page as best source to learn some content over even my university publications itself. I don't know whether I'll use this in future or not but I'm definitely going to use this for today. This page has some good content in it thank you 😊. Well doing great job 👏👏

[@ResearchByDesign]

Excellent...glad that I could help you with your studies. Wish you the best

[@pamellaoguadinma833]

This is more than what I ever thought I was going to get. God bless you real good

[@olofinsankolawole4894]

Thank you Dr. for this easy concise video.

[@ResearchByDesign]

Glad it was helpful!

[@JaenalynSimmons]

Hi! I love the video. Would love if you linked the other videos mentioned in the descriptions. I'm having trouble finding the video illustrating the Welch's and Games-Howell Post Hoc Interpretation.

[@ResearchByDesign]

Hi Jaenalyn, unfortunately the Welch video never got made (bummer), but I do have a new version in the works with JASP and SPSS. When I get that one up, I will link to it. Thanks for watching

[@JaenalynSimmons]

@@ResearchByDesign Appreciate it, thank you!

[@eyongmodest2062]

You are the best professor

[@frozenheart_07]

Nice fun example! This tutorial has helped me get an idea of what I'll need to do for my data. Thank you for sharing :D

[@ResearchByDesign]

Excellent! Very glad to hear that you have a clearer path forward with your data, now.

[@TheBusinessWorld9]

Amazing lecture🙏🙏🙏

[@nuramohammed480]

Thank you, such a helpful tutorial. Trying to do my stats tests for my Biomed dissertation!

[@ResearchByDesign]

Glad it was helpful!

[@emybacus3904]

very insightful and helpful in studying statistics. Thank you so much!

[@ResearchByDesign]

You're very welcome!

[@rizwanaqamar3270]

Awsome,and very comprehensive video

[@ResearchByDesign]

Glad it helped

[@pikiedust2907]

you do it so easily... I'm stuck with stats for a long time because I don't know what statistical analysis to do using SPSS for behavioural study.

[@ResearchByDesign]

I have a doc called "Hypothesis Testing" in the google drive for the course. IT has a section on how to choose a statistical test. It might be helpful: drive.google.com/drive/folders/1n9aCsq5j4dQ6m_sv62ohDI69aol3rW6Q?usp=sharing

[@char-5653]

Love this concept. Helped me a lot. Thanks!

[@ResearchByDesign]

You are welcome! And thank you for watching and commenting.

[@madhusmitaojha3632]

Thank you so much for your response sir.

[@marvins1373]

So which video is the one explaining the welch/games howell situation further? I did not find it. Otherwise, the content is really useful - thanks!

[@lichen938]

it's quiet useful. thank you. so clear

[@ResearchByDesign]

You are welcome! Thanks for commenting.

[@ijeomaezeh7807]

Thank you so much. But I still need to know how to correlate in spss when there is questionnaire for attitude and scores for performance

[@ehteshamali2893]

I love your videos sir!!! :)

[@ramakrishnan3074]

Thank you Sir, just mind-blowing explaination..Would love to check out Ur other videos..🙏

[@ResearchByDesign]

Thanks and you are welcome. Plenty of other videos on the channel, so dive in! kzbin.info

[@dhruvgoyal2346]

you are seriously awesome. thank you for this!

[@ResearchByDesign]

Thanks so much...it is great to have confirmation that I am, in fact, seriously awesome :o) Keep watching!

[@famavevershima2006]

Thanks for making and availing these videos. I however want to ask a question ❓ You mentioned that if groups have equal scores, then we could use Tukey's HSD test. What if the groups have unequal scores, which post hoc testing is appropriate? Thank you Dr.

[@ResearchByDesign]

If the groups have equal SIZES, then we use Tukey's. If the group sizes are different, we use the Tukey-Kramer test. The good news is that SPSS, JASP, etc. will quietly substitute the Tukey-Kramer test when the group sizes are unequal. You are safe to select Tukey as a post hoc. FYI: When group sizes are equal, Tukey’s HSD and Tukey-Kramer test are identical

[@ajsauzunovic2269]

Thank you very much nice man

[@ResearchByDesign]

You're welcome! Thank you for your kind words

[@hassanmuhammadsabo4450]

Hello. First of all Thank very much for this wonderful explanation. Its very helpful, and I have a question. I run One way anova in SPSS with my data i have got very high levene test Sig. which is .999 but i also got high P-value almost .986. How is this posible?

[@sharlotte_dse]

very informative, but I want to know if the number of sample in each group is unequal, how can we tackle this, thxxxxx

[@ResearchByDesign]

This is a video about unequal sample sizes using the t test. Same logic applies to ANOVA: kzbin.info/www/bejne/haqlkKpqjbOEa5o

[@kossonouprunelle7576]

Please is it possible to use ANOVA for both independent and dependent categorical variables, to compare the means ?

[@ResearchByDesign]

Yes it is. That is called a "Mixed ANOVA" or a Repeated Measures ANOVA with Between Subjects Factor". You do an RM ANOVA but drop in your categorical variable as a factor. Good luck

[@ΓιώργοςΨύλλος-μ3η]

Great job! Thank you.

[@pharm_2]

Will you please post the link for the video explaining Games-Howell?

[@boybitsvictoria3193]

Damn that example was so GOOOD!!!!! :D

[@empaulstube6947]

What would be the difference if we will just do this analysis thru the dropdown option of GLM - Univariate Analysis instead of Compare Means - One Way ANOVA?

[@ResearchByDesign]

You could absolutely do this with the GLM option. The output gives you more detail & options, and is what I would use if I was doing an analysis for real. The technique in this video is the best way to learn about ANOVA before you move on to the more powerful GLM option.

[@CPTBacardi]

Really informative, thank you.

[@ResearchByDesign]

Glad it was helpful! Thank you for commenting

[@Sabagat]

I can't find anywhere what would the write look like if Levene's is less than .05 and the assumption of homogeneity has been violated. :'(

[@ResearchByDesign]

The assumption of homogeneity of variance was tested using Levene’s test, F(1,44) = 3.51, p = .002; indicating that the assumption of homogeneity of variance had been violated. The t test results were interpreted using Welch's t test for homogeneity not assumed. The Welches t test was statistically significant, t(44) =2.86, p = .031.

[@madhusmitaojha3632]

very helpful video but i have an doubt for the welch test can explain why we used this test and what are the assumptions. Thank you from india

[@ResearchByDesign]

Welch's test does not have the assumption of homogeneity of variance, so if that assumption is violated (from Levene's test) then Welch's gives you an alternative test. It is still desirable that you have equal sample sizes for each group. Hope that helps.

[@lakeesharma3222]

Thank you so much for this informative video...now I have no confusion #p

[@ResearchByDesign]

Excellent...glad to have helped clear up some confusion. Cheers

[@chandapinkney1562]

Thank you for this video.

[@ResearchByDesign]

Glad it was helpful!

[@saliknazir4768]

What if Levene's test is significant? How should we interpret the result? Can anyone please guide?

[@ResearchByDesign]

If Levene's is significant, then the variances of the groups are not equal. Here is a video about what to do: kzbin.info/www/bejne/ap7Odo2unN2AiKM

[@kaushaltripathi1]

Atleast provide the video link you mentioned in the starting of the video

[@maurocampos5590]

Thanks for this!

[@aaronjacobfox4612]

Godbless you sir your the best,.

[@doctorcosta]

If each region did not have same number of participants then what would be ideal to choose instead of Tukey?

[@ResearchByDesign]

Unequal sample sizes requires calculating sum of squares differently (called Tukey-Kramer). Lots of work if doing Doing it by hand . When using SPSS, it will calculate HSD with that alternate formula (best I can tell). Other options would be Scheffe or Games-Howell (if the variances are unequal, too)

[@samshulamry4809]

Thanks very much for the video

[@sanamasroor469]

well defined..thank you so much. but i want to know one aspect. what the meaning oh "-" dash in significance level and F value. then what the actual reporting.

[@ResearchByDesign]

Not sure which part of the video you are referring to. Could you mention a time code so I can look? Also, the write up for the ANOVA ("actual reporting") starts around 10:48. Good luck!

[@mouaz7016]

Thank you for the good explanation, I would like to see videos about repeated -Measure ANOVA and the Factorial ANOVA.could we expect that soon?

[@ResearchByDesign]

Both of those are excellent suggestions. I have plenty of information about both of those topics...now I just need the time to make the videos. I think that you will see those topics covered eventually, though

[@fawadyasin8648]

Thank you so much It helped a lot

[@ResearchByDesign]

Glad it helped!

[@mamingyao7044]

hello, for example, variable: colour of shoes one wish to buy: red=1, pink=2, blue=3, white=4 if i want to enter 1+2 in the data view, how could i do that? coz it seems that only one data is allowed to be entered thanks loads

[@ResearchByDesign]

Yes, you can only enter one number per column/variable. If you want to indicate that someone chose both 1 and 2, you need four variables, each coded 1 = yes, 0 = no for whether they bought each color of shoes; however, you cannot do an ANOVA that way. For ANOVA, observations must be independent. Participants can be in only one group (red, pink, blue, or white). If they are in multiple groups, then you should exclude them from analysis.

[@amyr420]

Thank you! I was wondering, how do you know if a Levene's test is ok or not? What number should I look out for to see if Levene's test is legit or not :)

[@ResearchByDesign]

Video about Levene's: kzbin.info/www/bejne/ap7Odo2unN2AiKM yes, it is a useful test. Interpretation: If the sample size is n < 30, interpret Levene’s test at p < .05 If the sample size is n > 30, interpret Levene’s test at p < .01 If the sample size is n > 100, interpret Levene’s test at p < .001

[@crampledpaper]

how can I apply this into this case? Impacts of Limited books in Academic performance (referring to their semestral grade) Send help pls

[@ResearchByDesign]

It is difficult to answer given the limited information in your question...to do a one-way ANOVA you need one scale (continuous) variable (probably grades on a scale of 1 - 100%) and a categorical valuable (probably type of book). Good luck.

[@crampledpaper]

@@ResearchByDesign our research title is THE IMPACTS OF LIMITED TEXTBOOKS IN NAVIGATIONAL WATCHKEEPING 03 IN THE ACADEMIC PERFORMANCE OF TVL GRADE 12 STUDENTS. our questionnaire contains 1. What is your semestral grade in navIgational watchkeeping 03? 2. What are the impacts of limited textbooks in the academic performance? 2.1 positive impact 2.2 negative impact 3. What is the relationship between academic performance and impacts of limited textbooks in nvigational watchkeeping 03?

[@crampledpaper]

How can i use the one way anova for that? Huhu thanks for tge response😥

[@christianbistayan8718]

Hello, I am comparing groups with different number of population, what would be the best post hoc test to administer?

[@Lana-uq3br]

Tukey Kramer is for groups with different number of population. In spss the Tukey test implements and automatically produce Tukey Kramer results I think. So just click Tukey.

[@christianbistayan8718]

@@Lana-uq3br Thank you very much!

[@alicoco3930]

Nice video. Keep it up

[@julianatay9199]

hi, may i know, if i have to compare pre and post test, do i still use one way ANOVA? i tried following the steps in the video, but the data generated seperates the pre and post test. i wanted to compare the difference of pre and post test of 3 different groups.

[@sleddy2014]

Thanks so much for your videos, they are incredibly helpful! I want to run an ANOVA but one of my groups is slightly smaller, could you advise on which post hoc tests to run? Or if this is OK to do.

[@sakshikabra156]

Hey, the dependent variable should be a continuous variable for ANOVA test, while in this example number of complaints is a discrete variable. Is that okay?

[@ResearchByDesign]

That is a very good observation. When I was searching for an ANOVA example, "number of complaints" seemed like a good choice, but you're right to observe that it does not really make sense to interview 5 Munchkins and get multiple complaints from each ( better example would be to interview 5 fraternity members and ask how many drinks each had at a party). If we want to count something, we should count the number of wicked witch spell-castings in each region and maybe do a chi-square. I thought about that later, but I had already put too much work into the example. So good catch on the research design flaw, but the how-to and the math (and the animations) are all solid.

[@mariwanpsycho6490]

Well done

[@kyliebutler9374]

Hi all. Can someone please tell me in which video I can find more information about ANOVA with Welch's test (how to interpret it etc). It's mentioned in this video that this topic is dealt with in another video but I can't find it. Cheers.

[@ResearchByDesign]

Hi Kylie - thanks for the question about Welch's ANOVA. That video is still in production, along with one about contrasts in ANOVA. Will be ready for fall semester 2017. You interpret Welch's the same way (p < .05) but use the Games-Howell post-hoc. Send me an email (search Missouri State U) and I'll send you some notes.

[@dankindstrand]

@@ResearchByDesign Hi, im experiencing the same issue as the OP where my Levines test is significant - could you plese refer me to the video about how to interpret the Welch's test and the Games-Howell post-hoc. Thanks! Really great video!

[@LutherTaylor]

Do you have any recommendations for a good video guide on writing up a lab report? thank you for your uploads btw. Im currently getting screened for dsylexia and seems i have some memory issues that makes it hard for me to retain information. I had a fair amount of difficulty working through anova but it was still sooo much simplier than what my university had made. respect! p.s, is it you that is doing all the art and motion graphics? :D

[@abhishekasthana1911]

Please help us with a video to apply ANOVA on a Likert scale

[@ResearchByDesign]

Great idea. The actual ANOVA would be conducted the same way that you see in this video. Using a Likert scale, you will typically combine multiple items into a single score (using a mean or sum total). That single score is your DV and your IV would the the grouping with 3 levels.

[@worldofinformation815]

great, thank you

[@81818alia]

I want to compare 56 groups. Running ANOVA on SPSS, I receive a warning that Post hoc tests cannot be performed as I have more than 50 groups. Can you kindly suggest other options?

[@ResearchByDesign]

Your best option is to simplify and scale down the number of groups. Unless you have reason to believe that every group could be meaningfully distinct from all of the others, the groups probably have similarities or sub-groupings. Consider combining related groups and/or doing planned contrasts (a priori) rather than post hoc tests.

[@tbear654]

"...one group has fewer than two cases" What does this mean?

[@sychomadman]

Thank you sooo muchhh!!

[@ResearchByDesign]

Glad that it was helpful! Thank you for watching.

[@maramezina1094]

Could you show us how to download SPSS please?

[@aaronjacobfox4612]

you go to a computer shop

[@RahulKumar-mp3jq]

PLEASE PROVIDE ME THE VIDEO LINK FOR GAMES - HOWELL POST HOC

[@ResearchByDesign]

Yeah, about that...that G-H is a video for which I have a script, but have not recorded it yet. The bottom line is that you (a) check for violations of homogeneity of variance (Levene's test), then if the assumption is violated (b) interpret the post hoc using Games-Howell, instead. This video (12-6) shows how to run Games-Howell, including the Leven's test, so you could follow these steps exactly and then use G-H if Levene's is significant.

[@81818alia]

Thanks a million

[@mikehanzmariano790]

Warnings There are fewer than two groups for dependent variable math. No statistics are computed. this pops up?

[@ResearchByDesign]

I would only be able to see for sure if I saw the data set but I suspect that one of your variables is a constant (i.e. all of the values are the same). SPSS is looking for numbers that indicate groups, it needs at least two, and it found less than 2. Check the data and make sure your grouping variable has 2 or more groups. Good luck.

[@benysmart1643]

Thank you

[@ResearchByDesign]

You're welcome

[@sagna2793]

Hello sir can I get your contact info please

[@chanceroberts7494]

My stat class is gonna be the end of me

[@missqueenwendy]

I hope you survived. Going back to school at 44 and I'm in my last semester with Research Methods II. I had to up my anxiety med. (Whew!)

[@katerina6297]

You sound like Sheldon

[@maramezina1094]

Very useful thank you

[@inasalameed4629]

thank you