How to Use SPSS: Choosing the Appropriate Statistical Test

A step-by-step approach for choosing an appropriate statistcal test for data analysis.

[@emt9103]

There are all sorts of materials on the internet regarding the application of the statistical tests (T-test, ANOVA, Regressions etc). But there is hardly any video about which test should be chosen and why. This video perfectly fills the gap in this area. Five stars!

[@lsobrien]

You, sir, are a scholar and a gentleman. Thank you for this video.

[@SolomonEdwardsojiambo]

This is the best lecture anyone can ever get. You make stats easy and lovely. Thank you.

[@50Synapse]

Very lucid and practical review on how to choose the appropriate test to interpret results. Better yet, the process helps crystallize how to design the appropriate research protocol. Thank you.

[@meddiekalumba26]

Very clearly thought out presentation. Thanks so much.

[@Ultramama2017]

I think this video is really helpful for those who just started learning statistics, and to the research newbie who's in the planning stage of research. Thank you! ^^

[@djurosvemirac88]

Great video. Finally someone explained everything in a very simple way. Tnx a lot!

[@janedigiovanni8284]

Thank you so much! This video was just what I needed. After two semesters of statistics, I just needed the overview. Now, I can see the big picture. Now, after feeling confused for a very long time, I get it. Yeah! Thank you, thank you, thank you.

[@evaadolfo2765]

I am actually on the process of choosing the test statistics to use...and God guided me to this video. Awesomeness!!! Thank you for your wisdom. More power to you. Please keep making very good videos.

[@gmaxtrackstars1]

I was able to follow this video to the end and not get lost or confused. It helped me so much.... Thank you

[@pedro-ivannavarronaranjo6682]

Excellent class and practical use of different statistical tests. Amazing

[@gururajp2006]

Thank you so much sir, it's short, simple and sweet!

[@andrewisaacs3435]

I have found this video extremely helpful. Thanks a million and more times.

[@sisayabayneh8442]

Really thank you so much for this very clear, comprehensive and greatly helpful lecture. I am very new to statistical analysis but with very deep interest to learn it and scar that need rehabilitated and cured. I have get with you the solution. Thanks, Sisay -Ethiopia, Africa

[@BW-kv3fm]

Organized and clear. thanks a lot!

[@robertarnott3103]

The Best Video on this topic I have ever seen, there is nothing better around, thanking you Sir

[@kuutolkwamepeter3075]

Wonderful! Thanks for your impact my doubts have been cleared.

[@daniellemartin3667]

This has been really helpful, with all the books on SPSS, it doesn't explain it as well as this does, thank you for uploading !!

[@jameswu1768]

2 Thumbs up! Helped me tremendously to grasp the idea! Thank you!

[@LaSonyaThompson]

I'm grateful that the angels of Heaven guided me to this video for a breakthrough in Statistics!

[@emeraldkatt4899]

Very clear and concise explanation! Thank you very much for sharing!

[@passion830217]

It is a fabulous video to review what i have learned from statistic 1 and statistic 2 classes. You are the best! thanks!

[@andrecurtis2335]

This video is a gold mine!!! Thank you a million. It's taken almost an hour to get through 10 minutes. I can't stop taking notes...😁😁

[@tacappaert]

So glad to hear that!

[@citybranding5517]

Thank you this is very helpful. It is very well explained and the examples are clear and clarify a complex subject very well.

[@swapnadeshpande5853]

very helpful! Thank you for uploading!

[@mikan7532]

THANK YOU THANK YOU SIR. YOURE BETTER THAN MY LECTURER ^.^ !! I enjoyed watching this video that solves my doubts ^.^

[@brohusseinable]

Absolute excellence!

[@mcqueenbuenavides6940]

Thank you for the video.

[@MohdFarhan-ye9kc]

I never understood these tests, or how to use them, but this video made it so easy to understand. Thank you.

[@tacappaert]

You're very welcome!

[@mathewsmonde7917]

Thank you i found it very helpful

[@4444saso]

so thanks for very helpful information

[@rahmanahamedyt]

Thanks a lot for this useful video demonstration

[@ashrafothman7654]

Thank you,well done and very helpful

[@clintonarum6935]

a wonderful video, thank you man

[@TheRealAK786]

Thank you, from London!

[@fahdb.altherwi1566]

Marvelously marvelous ... A THOUSAND OF THANKS

[@qianliu3852]

addressed a lot of confusion by one video!!

[@pozitivpszichologus]

Its a great video, but i am still not sure about what test to use if i have a binominal dependent variable, and i have nominal and scale independent variables too. could you help me? thanks in advance

[@rochellerobinson-levant8835]

Fantastic clarity. Really well done.

[@tacappaert]

Thank you kindly!

[@debrahisaiah7025]

Nice Tutorials. thanks

[@yishiliu4156]

Thanks, the best statistical aid ever！

[@draganabozic2317]

Excellent video! Do you have one on testing normality (skewness, kurtosis, Shapiro-Wilk test) and reporting results?

[@tacappaert]

+Dragana Bozic Yes, kzbin.info/www/bejne/mWOxf5SLaphqf9E

[@riazmarakkar]

really helpful content thanks

[@gordondraskic4348]

Thanks for this great video! Anyway, I still have a question (example) which I didn't find in the video. Which test to use when you have more than one independent variables (for example, locality and sex) and more than two dependent variables? Is the Two-Way MANOVA right choice or it should be something else? Thanks

[@tacappaert]

As long as all assumptions are met, then the Two-Way MANOVA would be appropriate.

[@cuteMuso]

Thank you so much

[@pavipavatana5972]

Thanks!!!

[@hamidkhaleghi]

Good job! Thanks for making such a helpful video. It answered a lot of my questions. Except I am still wondering when to use t test or f test or Z test or Wald test! Many thanks. H.

[@tacappaert]

+hamid khaleghi A z test is used when comparing a sample to it's population, a t-test when comparing two independent or paired groups and an F test when comparing 3 or more independent or dependent groups.

[@vegardarnesen8801]

Thank you so much for such a great video! I have one question, and I hope you might have an answer to that. I´m doing my masters in public health nutrition and using statistics for nye research question. My question is: Does people who drink soda more than x-times per week (categorical variable) have higher blood pressure (continuous) than those who don¨t drink soda more than x-times per week. I think I should use independent t-test for the main question, but I also do have cut-off point for my blood pressure (good or bad value). Is t-test, and maybe ANOVA appropriate for this question, or am I way out there?:) Thank you so much for you help. Masterstudent in Norway.

[@tacappaert]

If you keep your soda consumption variable as a two level category and the outcome numeric then the independent t-test would be appropriate. If you change your outcome to a category, then you could examine the influence of category of soda consumption using a chi-square test (2x2 table).

[@mohamedabdallah26]

full thanks

[@dhanyasrinath26]

Hey, This was helpful but I'm still a little unsure of what test to use for my research. I have 4 scenarios, 4 examples of advertisements and I have the same set of questions for each scenario. I need to prove which advertisement was most effective based on the results. Which test can I use to test the effectiveness of each scenario and compare to see which did best? I have to test which scenario has best customer loyalty basically. how do i check that? Would appreciate any insights from you! Thank you so much in advance

[@tacappaert]

+Dhanya Srinath If different groups of people are exposed to each of your 4 scenarios then an ANOVA would be appropriate.

[@chongtzeweai7306]

Thanks for this great video! I am currently doing research on internal audit function, what test should i use if there are four dependent variables and one independent variable? Is it MANOVA? I am using 5-point likert scale to set the questionnaire. I hope you can help please. Thanks!

[@tacappaert]

Chong Tze Weai Yes, you could use MANOVA. However, likert scale data is often skewed, so you will have check to make sure the data meets all the assumptions of MANOVA.

[@missbexychick]

Hi there, your vids are excellent. Just wondering if you could help me with which type of ANOVA to use. I have four different mood conditions (sad, happy etc.). I am measuring individuals moods at time 1 and am measuring their moods at time 2 after the mood inductions (to see if the mood inductions have worked). The mood measures are questionnaires with four questions on their mood. Is this a two way repeated measures ANOVA or one way repeated measures ANOVA? How would I do this on SPSS?

[@tacappaert]

missbexychick It would be a two way repeated measures ANOVA. Also known as a mixed design or split plot ANOVA. Check out my ANOVA playlist for a video demonstrating this technique.

[@debrahisaiah7025]

thanks a lot

[@marwatalbi7125]

i am using the mann-whitney U test in my study. should i look for the median of my ordial variable (dependent variable) before running the test or simply use the test as it is with the answers of the repondents ?

[@tacappaert]

+Marwa Talbi Just use the test but when you describe your groups, use the median score.

[@jenniferhrastar6919]

Hi thanks for the really great video. So I figured to do a dependent t-test to see if there exists any significant difference, if assumptions are met. However how can I classify the difference of the two variables incorporated in my hypothesis: A is better rated than B?

[@tacappaert]

+Jennifer Hrastar It will be based upon the group means as to which group has the "better" score.

[@avaphillips9848]

Hi I wondered if you could help me chose the correct tests. I have one IV which has three levels (conditions) and participants are doing the SAME test twice, once at beginning then once again five years later. Im looking at the difference between three different teaching styles and their affect on how much they decrease symptoms of a specific mental disorder. Am i right in think that it would be a 2x3 way anova, followed by post hoc test to further investigate the interaction effect, then an independent samples test? Would be real grateful for your advice! THankyou

[@tacappaert]

Ava Phillips You would perform a two-way repeated measures ANOVA with treatment group as one factor and time as the second factor.

[@nirmeeta]

Your explanation is very precise and has helped me to get better understanding but for a person like me whose native language is not english have to depend on subtitles. I bet this is not you who have written the script ... I suggest u to go through it once before posting ... Laughter guaranteed :D

[@emmamarjan2411]

What about if you just have questions answered using a likert scale? I don't have any variables/i'm not comparing any groups or anything, just simple likert scale questions. Which test would I use?

[@tacappaert]

+Lovery Culture You could use basic descriptive measures like frequency, percentage or a median score.

[@solentebaptiste112]

Hello. I am comparing 6 faculties with perception of healthy eating and gender with healthy eating. I have used Anova and Ttest however i was wondering if I was supposed to use CHIsquare test for a question involving GROUPS. eg. How many portions of Veg you ate? 1-2 por, 2-3 por, 3-4 por, 4-5 por (between Faculty comparison (6 groups) and Gender comparison(2 groups.)

[@tacappaert]

+Solente Baptiste If your outcome is in categories as you described then chi-square is more appropriate.

[@joshsnell9666]

Hi quick question, I want to know whether scientific background (3 groups: science, intermediate and non science) has a significant effect on scores on my measure. What test do you recommend?

[@tacappaert]

Josh snell If your outcome measure is numeric, then a one-way ANOVA would be appropriate.

[@joshsnell9666]

Okay, thanks a lot for your help

[@wimbayichasaya.1068]

great video, thanx for posting. how do you analyse a question like this one : Identify the level of employee motivation?' using SPSS. Lets say the data was collected using a 5 point Likert scale

[@tacappaert]

+Wimbayi Chasaya You could identify the mean or median for your measure. If you wanted to compare two groups on their motivation, then a Mann-Whitney test could be used.

[@sitotagetachew4392]

Stay Blessed bro!

[@tacappaert]

You too

[@walaahatahet783]

what test would i use to measure compliance of healthcare workers with a set of policies or one policy???

[@tacappaert]

+Walaa Hatahet If your measure of compliance is numeric you could use an independent t-test. If it is nominal, you could use a chi-square.

[@blimeyitsjessie]

I have 2 independent variables (silent reading versus oral reading) and my dependent variable is memory recall. My research question is will silent reading or oral reading result in better memory recall. It's a within subjects test. 1 group tested on oral reading and then again on silent reading. I'm thinking I'd use an Anova but i'm not positive.

[@tacappaert]

+Jessica Marie If the same group of subjects was tested twice, once with silent reading and then again with oral reading, then a dependent/paired t-test will be appropriate.

[@gwenward56]

This is helpful but I am trying to figure out which statistic test to use if you just have the frequency.

[@tacappaert]

Gwen Ward If the frequency is numeric (how many times have you read a book this week?) then t-tests or ANOVAs can be used. If the frequency is a category (how many people said yes vs. how many people said no), then you would be able to use a chi-square analysis.

[@chuksnnamdimedoh2930]

Wow! A great video. You are such a scholar Sir, More Grace. Am still a little confuse what test to use though, am working on a dissertation: Impact of automation/manual process on Record management. Am using five categories to measure the impact , and each category has four different sub categories( with an agree or disagree option). Could help Sir. Thanks in Advance

[@tacappaert]

Chuks Nnamdi Medoh What is your outcome? Is it numeric or categorical? If it is numeric then you could use t-tests or ANOVAs. If it is categorical, then Chi-Square would be more appropriate.

[@GiselleMarieXIII]

ty

[@JoyJoy-jv7ss]

Marvelous inf

[@udaybhatnagar18]

How do I measure Consumer purchase decision (dependant variable) ?

[@tacappaert]

+Uday Bhatnagar Do you want to compare purchase decision between two or more groups? If your dependent variable was categorical, then a chi-square would be most appropriate.

[@marwatalbi7125]

it is a great video. but i face one problem in my thesis. i have to use a non parametric test. but i am confused betwenn Mann Whitney U test or the Chi square test. my research question is the following: does gender influence students' intercultural competence (ICC). given that to measure ICC, a 5point likert scale was used. can you help me please

[@tacappaert]

+Marwa Talbi I would use Mann-Whitney due to the ordinal nature of your outcome.

[@marwatalbi7125]

+TheRMUoHP Biostatistics Resource Channel thank you

[@tacappaert]

Kali Apostol For the numeric outcome you would use a independent t-test and for the nominal outcome you would use a Chi-Square.

[@drcycy]

+TheRMUoHP Biostatistics Resource Channel thanks for your lovely video. i am doing a research on online banking acceptance. I used quantitative methods and collected data from questionnaires. it was answered by users and non users. i am confused on what method house. i started using the chi square to test my hypothesis but i am lost ... i am not sure how to pic out a method. THANKS

[@isaacakubo9305]

how can i view your youtube online

[@davutatalay1142]

is it possible to obtain this presentation's slides?

[@tacappaert]

+Davut Atalay Message me your email address and I would be happy to send them to you.

[@SalmanAli-og4oe]

It would be great if you could share the presentation slides too. Thanks

[@tacappaert]

Salman Ali Sure, send me a message with your email address and I will send it along.

[@SalmanAli-og4oe]

***** Thank you! That will be great. Please send it to [e-mail]

[@kibkabwondossen42]

Hello can i get the ppt data

[@gwenward56]

A political pollster is curious about the effects of a town hall meeting on people's intentions to support a state proposition that would legalize gambling. He interviews people as they leave the meeting. He records these frequencies in the table below. What is the appropriate test to analyze the data. Less likely to support 25 , No change 12, more likely to support 9

[@gwenward56]

Gwen Ward Can you help me please

[@tacappaert]

Gwen Ward You would use the Chi Square Test for Goodness of Fit.