What a clear, effective and efficient explanation of OLR analysis. Great job! 👍🏼 This lecture gives the confidence to start working on my assignment today. Thank you so much Dr for your help.
WOW! EXCELLENT!!! VERY CLEAR, DIRECT, EFFECTIVE WAY OF TEACHING. EXCELLENT PACE! The dullest student will excel under your mentorship. Simply over and beyond!! This is what teaching is. Hope others who come here and RUSH things through thinking they have imparted knowledge can learn from you!!! THANK YOU SO MUCH!
Dear Dr. Latif, thank you so much for this incredible lecture! It has helped me a lot in understanding what ordinal regression is. I am now trying to implement it for the analysis of extremely negatively skewed continuous dependent variable (the scale varies from 0 to 1, and there are a lot of values that equal 1). I have split this scale into 5 equal categories (0.0-0.2, 0.2-0.4…), included them into a regression model (together with predictor variables). Could you please tell me if it is a correct way to split such data? Another question if I may ask, how does the number of categories of the dependent variable affect the analysis? And lastly, is it important to have P values
Pleasure. I am glad you liked the data. If the data is not normal, why not use Bootstrapping procedure in regression analysis. Otherwise you can transform the data as you have. Not all categories have to be significant.
Thank you for this video! You are the first one to explain how to get the odds ratio!! Actually, I would like to have your opinion. I am doing ordinal regression for my bachelor thesis and all my variables are ordinal. Since I asked around 4 questions to assess each variable, I had 4 values for each variable so I took the average. However, now my values are not 1,2,3,4 and 5 anymore but now have decimals. How would you use this data then? Should I group them or treat them as continous and put them into the Covariates box in SPSS? Any suggestions? Help is greatly appreciated
Hi Professor, thank you for the video, I had a question about the significance of the P-values in the parameter estimates: is this like ANOVA: between the levels of each categorical predictor that there is no difference when comparing the odds ratio for both levels?
A great video thank you. Could you explain what the error message at the start of your results means for the interpretation? I get a similar error message that a percentage of my cells have zero frequencies. Does this mean there was little relationship between my dependent variable and my predictors?
Very informative video.. I wanted to find out if i can use ordinary logistic regression for a non normally distributed data ..or what is the right regression analysis to use for non normlly distributed data. Also if the siginificance level in your model fitting information is >¤ do you still have to use the model..
Thank you so much for this video! Do you also have a video about how to report the results? I only found a video about regression analysis but not about ordinal. Furthermore, what do I do if the Goodness-of-Fit for Pearson is
Thank you sir for the valuable knowledge you share. This really helped me. However, I'm just wondering about the p-value. How can we still interpret the estimates coefficients and the odds ratios while the p-value is higher than alpha (e.g. 0.05)? It happened in my case and I saw it also happened in your video. I cannot get the information on those two when I do not have sufficient evidence that the variable or the level does not significantly affect odds of the outcome. Can you help me sir please.
I am glad you liked it. Not every relationship is significant. If the impact is insignificant, you will have to provide justification as to the possible reasons for the insignificant results. This video may also help ua-cam.com/video/xPZymvNG-gk/v-deo.html
Dear professor, this was such an incredibly well-done video, thank you very much. I have a statistical case which has been causing me a dilemma. Would it be possible to give you the context, and you could kindly direct me to which test I should be reading about? If yes, should I post the context of the data I have here, or somewhere else? Again, thank you very much.
@@researchwithfawad Thank you so much Professor! My case is as follows: I have a set of demographic variables just like the one you used. My other set: 1-Contains multiple variables, each one representing a photo. 2-The photos go from most unnatural to most natural. 3-We asked every participants to rate each photo on a Likert scale (1-5) that went from most unnatural to most natural. 4-Every participant got half the photos picked randomly, not all, and in random order. 5-I want to know which photo (so which variable, as a whole) acts as the cut-off for every demographic group. For example: Photo #3 is the natural Cut-off for middle age group, photo #5 is the cut-off for high salary group, and so on.
Thanks for your interest. You will need to take the sum of the individual items. Let say, i have Organizational Commitment measured using 4 items COM1, COM2, COM3, COM4 If you have it in SPSS, Go to Transform -> Compute Variable In the Target Variable Enter the Name of the New Variable that is to be created based on taking the average, let say COMM. In the numeric expression type in Mean(COM1, COM2, COM3, COM4) Press OK. The new variable is created at the end of the Data View and is also visit in the variable view. You have now composite score for each respondent that you can use in regression.
How can we do ordinal regression where the dependent variable and independent variables have large number of items whose observation is collected in likert scale? can we use the transformed mean for analysis of ordinal regression? my research has 1 dependent variable with 22 items and 3 independent variables each with 24 items. all the items are likert scale type
Hi, Prof. I'm wondering if we can use ordinal logistic regression if our DV is Likert scale (but we calculated the mean score according to the domain)? For example, the physical activities domain has 3 items with 4-point Likert scales. If the respondents answer 3,1,4 to each item, the total score will be 8 and the mean score will be 2.66. Is this still considered appropriate for an ordinal regression analysis?
thank you for your video. However, may I know what should I do to handle the redundant parameter in this case? actually I don't even know what 'redundant parameter' means and how it is gonna affect my research result
A redundant parameter is one that doesn't impact the outcome. I do not recommend its removal rather i recommend you to explain what a certain factor may be insignificant in the context of your study.
The regression procedures for categorical dependent variables do not have collinearity diagnostics. However, you can use the linear Regression procedure for this purpose. Collinearity statistics in regression concern the relationships among the predictors, ignoring the dependent variable. So, you can run REGRESSION with the same list of predictors and dependent variable as you wish to use in LOGISTIC REGRESSION (for example) and request the collinearity diagnostics.
Hi Fawad, just got your nice message. Would you be able to tell me whether or not I have to convert my likert scale IVs to present and absolute values before running the ordinal logistic analysis (I read that somewhere). thanks
Hi Fawad, huge thanks for this video! I got one question, if i run my analyses i get the warning: "there are 118(66,7%) cells (i.e. dependent variable levels by observed combinations of predictor variable values) with zero frequencies. I checked the assumptions and found no values that are not being used. Can you explain if this is a problem, what's causing the problem and mostly how i can fix it if is a problem... Thank you very much again!
Can I use continuous variables in Ordinal Logistic Regression? I tried using it, but it seems like continuous variables are converted to nominal variables when looking at SPSS results. It means that when using a continuous variable in the model, I obtain too many parameters corresponding to each different value in this continuous variable while I expected there to be only one parameter corresponding to that variable.
Thanks a lot! Yet one clarification...In your depenedent variable you have 5 star rating there. yet only four's output was only there. what about the 5th one. IF the 5th one won't come in output; pls explain why?, Thanks in advance! :)
Hi sir. I have a question. In statistics non parametric tests are the ones used for ordinal data, and I've never really heard of central limit theorem applied to ordinal data. How can we theoretically justify the use of Wald test in ordinal regression, since it's a parametric test?
In statistics, nonparametric tests are commonly used when working with ordinal data. Parametric tests, such as the Wald test, make assumptions about the underlying distribution of the data. The Wald test is a parametric test commonly used in regression analysis, including ordinal regression, which assumes that the dependent variable follows a particular distribution (e.g., a logistic distribution in the case of ordinal regression). The theoretical justification for using the Wald test in ordinal regression lies in the asymptotic properties of the estimator. The Wald test relies on the central limit theorem, which states that under certain conditions, the distribution of the estimator becomes approximately normal as the sample size increases. In the case of ordinal regression, the Wald test is based on maximum likelihood estimation, which has been shown to have asymptotic normality properties.
@@researchwithfawad You are such a big help for my research! May Allah bless you! by the way is the result okay if the all parameter estimates are non significant? Thank for answering
Thanks. If you are looking for model development, you may watch the video in the attached playlist. How to Find Research Gaps/New Topic for Research: ua-cam.com/play/PLb7vm6tsQ3KslXSw1BOb3Q0yJnksLyi2Z.html
Hi sir, Is the data "interest" in ordinal or scale? What happens if the data interest is ordinal, because the dependent variable should be ordinal so it uses ordinal logit regression
Sir,can we apply OLS regression in determine impact of demographics (Age ,Income,gender, educarion)on dependent variable financial management behavior (Likert scale)...do we need to convert likert scale
If there a multiple items in the DV, you will need to take its mean and if the predictor categorical variables have more than two categories you will need to create dummy variables and use linear regression.
@@researchwithfawad Actually the sample size is small (90 observations) but I changed IVs from factors to covariates and now it seems much more clear although most effects are not significant
@@researchwithfawad My DV is ordinal, a Likert scale (not motivated to highly motivated) how do i know if male (or for married) for example are moderately motivated or fairly motivated)? And whats the difference between parameter estimates and log of odd. thanks Its urgent huhu thanks for answering! God bless
I have a question sir, what if one has an Independent Variable with 5 categories, ie, strongly disagree, disagree, neutral, agree and strongly agree. assuming strongly agree is set as the baseline/reference category, how can you interpret the estimate value if; strongly disagree is -ve sign, disagree is -ve sign, neutral is +ve sign, and agree is -ve sign while baseline is redundant (0)？and in another case strongly disagree is -ve sign, disagree is -ve sign, neutral is +ve sign, and agree is +ve sign while baseline is redundant (0)?
A positive or negative sign with the estimate in the Parameter Estimates table shows the likelihood of falling into a particular category. + sign is associated with an increase likelihood of case falling into a higher category in the dependent variable. - sign is associated with an increase likelihood of case falling into a lower category in the dependent variable. For instance, a Positive Sign with and IV with 5 Categories would mean that with increase in the value of that variable, there is a higher chance of falling into a higher category on the DV.
@@researchwithfawad Thank you for the response sir, very much appreciated. Perhaps my question was not clear enough, what I was intending to ask is if the variable has different signs at the same time at different categories. For instance a construct “happiness” with five categories and some categories have - sign while some have + sign in the estimates. How would one interpret happiness? When it has some categories that are + and some that are - in the estimate at the same time.
Hi professor, in this video the DV is ordinal, but running the test in SPSS you have taken DV as ratio data instead of ordinal data, is this correct pls clarify, thank you in advance
At last a video for understanding and comprehending the entire process the easiest way! Thank you so much! Godspeed! ❤
I cannot thank you enough for this video! I wish my professor taught like this.
Pleasure. I am glad you liked it. Please do share in your research circle.
THE EASIEST AND UNDERSTANDABLE VIDEO IV EVER WATCH SOO FAR IN MY 4 YEARS OF STUDY. THANKYOU SOO MUCH !!
Pleasure. I am glad you liked it.
What a clear, effective and efficient explanation of OLR analysis. Great job! 👍🏼 This lecture gives the confidence to start working on my assignment
today.
Thank you so much Dr for your help.
I am glad you liked it.
Excellently and beautifully rendered. It makes a lot of sense now. Thank you for this resource
Thanks. I am glad you liked it.
WOW! EXCELLENT!!! VERY CLEAR, DIRECT, EFFECTIVE WAY OF TEACHING. EXCELLENT PACE! The dullest student will excel under your mentorship. Simply over and beyond!! This is what teaching is. Hope others who come here and RUSH things through thinking they have imparted knowledge can learn from you!!!
THANK YOU SO MUCH!
You are most welcome. Do Share in your research circle.
I was really struggling to understand how to interpret my results before this video. You explained everything so well - thank you so much!
You're very welcome!
Thank you for posting this informative breakdown!
Glad it was helpful!
Thank you Prof. This is so detailed. Thank you once again.
Thanks. I am glad you liked it.
Professor shukran! You saved me! Amazing explanation and very easy to follow.
You are welcome! I am glad you liked the video.
Awesome video! Great explanation, thank you so much :)
Glad it was helpful!
Thank you Fawad for your excellent presentation and it helps me a lot!
You are very welcome
Brilliant stuff, Fawad.
This is really a great presentation to give beginner confidence to go with OLR
Thank you! I am glad you liked it.
Great explanation. I love this
Thanks for watching.
Thank you for this well-detailed explanation. ❤
Glad it was helpful!
Very useful, u really save my life !!!!!! Power impact for carry out my sass assignment
Great to hear!
Great video, helped a lot!
Glad it helped!
Good Job ! Jazaakallah Khairan!
From#Ethiopia
Thanks. I m glad you liked it.
thank you so much, this is so clear
Thanks. I am glad you liked it.
Fawad - I have looked in a number of places for a better description....and I can't find one. Many thanks.
You're welcome
This video helped me alot .. 😊thankyou ❤❤
Thanks. I am glad you liked it.
Thank you very much!
You're welcome!
Thank you for this
My pleasure!
Dear Dr. Latif, thank you so much for this incredible lecture! It has helped me a lot in understanding what ordinal regression is. I am now trying to implement it for the analysis of extremely negatively skewed continuous dependent variable (the scale varies from 0 to 1, and there are a lot of values that equal 1). I have split this scale into 5 equal categories (0.0-0.2, 0.2-0.4…), included them into a regression model (together with predictor variables). Could you please tell me if it is a correct way to split such data? Another question if I may ask, how does the number of categories of the dependent variable affect the analysis? And lastly, is it important to have P values
Pleasure. I am glad you liked the data. If the data is not normal, why not use Bootstrapping procedure in regression analysis. Otherwise you can transform the data as you have. Not all categories have to be significant.
Thank u so much ser
Most welcome
Thank you prof. I want to ask one thing. Have you done an example on the eliminating by the 10% ratio difference calculation? Thank you very much.
Not yet!
excellent sir
Many many thanks
Thank you for this video! You are the first one to explain how to get the odds ratio!!
Actually, I would like to have your opinion. I am doing ordinal regression for my bachelor thesis and all my variables are ordinal. Since I asked around 4 questions to assess each variable, I had 4 values for each variable so I took the average. However, now my values are not 1,2,3,4 and 5 anymore but now have decimals. How would you use this data then? Should I group them or treat them as continous and put them into the Covariates box in SPSS? Any suggestions? Help is greatly appreciated
Thanks for watching. Use Linear Regression.
Hi Professor, thank you for the video, I had a question about the significance of the P-values in the parameter estimates: is this like ANOVA: between the levels of each categorical predictor that there is no difference when comparing the odds ratio for both levels?
Yes, you may also interpret it like that.
Hello Professor, thank you for this valuable information, is it possible to also provide an example on how to analyze Multinomial Logistic Regression?
I hope i can soon make a video on that.
well explained
Thank you.
A great video thank you. Could you explain what the error message at the start of your results means for the interpretation? I get a similar error message that a percentage of my cells have zero frequencies. Does this mean there was little relationship between my dependent variable and my predictors?
Thanks for watching. In some of the categories there is no data.
Very informative video..
I wanted to find out if i can use ordinary logistic regression for a non normally distributed data ..or what is the right regression analysis to use for non normlly distributed data.
Also if the siginificance level in your model fitting information is >¤ do you still have to use the model..
Normally, the regression technique is robust to minor violations of normality. Please check for skewness and kurtosis.
Thanks a lot Prof. But I get slightly confused with the last sentence in your video. Do you mean an insignificant test?
If p value is over 0.05 it is insignificant.
Thank you so much for this video! Do you also have a video about how to report the results? I only found a video about regression analysis but not about ordinal. Furthermore, what do I do if the Goodness-of-Fit for Pearson is
Thanks for watching. No nothing yet.
Thank you sir for the valuable knowledge you share. This really helped me. However, I'm just wondering about the p-value. How can we still interpret the estimates coefficients and the odds ratios while the p-value is higher than alpha (e.g. 0.05)? It happened in my case and I saw it also happened in your video. I cannot get the information on those two when I do not have sufficient evidence that the variable or the level does not significantly affect odds of the outcome. Can you help me sir please.
I am glad you liked it. Not every relationship is significant. If the impact is insignificant, you will have to provide justification as to the possible reasons for the insignificant results. This video may also help
ua-cam.com/video/xPZymvNG-gk/v-deo.html
Dear professor, this was such an incredibly well-done video, thank you very much. I have a statistical case which has been causing me a dilemma. Would it be possible to give you the context, and you could kindly direct me to which test I should be reading about?
If yes, should I post the context of the data I have here, or somewhere else?
Again, thank you very much.
Thanks for watching. You can write to me here.
@@researchwithfawad Thank you so much Professor!
My case is as follows:
I have a set of demographic variables just like the one you used.
My other set:
1-Contains multiple variables, each one representing a photo.
2-The photos go from most unnatural to most natural.
3-We asked every participants to rate each photo on a Likert scale (1-5) that went from most unnatural to most natural.
4-Every participant got half the photos picked randomly, not all, and in random order.
5-I want to know which photo (so which variable, as a whole) acts as the cut-off for every demographic group. For example: Photo #3 is the natural Cut-off for middle age group, photo #5 is the cut-off for high salary group, and so on.
I would like to see how you entered the data, can you share a sample on my email kh.fawad83@gmail.com
Thank you very much. But when we conclude about the parameter estimates won’t we consider statistical significance value??
Yes we will.
Thanks Sir,can you please tell me how to combine likerscales 5 values (showing one variable) into one for analySis
Thanks for your interest.
You will need to take the sum of the individual items. Let say, i have Organizational Commitment measured using 4 items COM1, COM2, COM3, COM4
If you have it in SPSS,
Go to Transform -> Compute Variable
In the Target Variable Enter the Name of the New Variable that is to be created based on taking the average, let say COMM.
In the numeric expression type in
Mean(COM1, COM2, COM3, COM4)
Press OK. The new variable is created at the end of the Data View and is also visit in the variable view. You have now composite score for each respondent that you can use in regression.
If you have an directed hypothesis, do you split the p value in half while interpreting? Does someone know?
You can if the hypothesis is directional.
How can we do ordinal regression where the dependent variable and independent variables have large number of items whose observation is collected in likert scale? can we use the transformed mean for analysis of ordinal regression? my research has 1 dependent variable with 22 items and 3 independent variables each with 24 items. all the items are likert scale type
In this case Ordinal Regression is not performed. You will perform linear regression.
Hi, Prof. I'm wondering if we can use ordinal logistic regression if our DV is Likert scale (but we calculated the mean score according to the domain)? For example, the physical activities domain has 3 items with 4-point Likert scales. If the respondents answer 3,1,4 to each item, the total score will be 8 and the mean score will be 2.66. Is this still considered appropriate for an ordinal regression analysis?
Do not use ordinal Logistic Regression when the DV is continuous as in your case with mean score.
@@researchwithfawad Which regression analysis should we use in above case (continuous)?
thank you for your video. However, may I know what should I do to handle the redundant parameter in this case? actually I don't even know what 'redundant parameter' means and how it is gonna affect my research result
A redundant parameter is one that doesn't impact the outcome. I do not recommend its removal rather i recommend you to explain what a certain factor may be insignificant in the context of your study.
Thanks a lot for this video sir..Why we are not using stress variable sir..
Thanks for watching. It is an example, there is no particular reason.
Does anyone know how to change the reference category for the categorical predictor variables?
Hi. For instance, you have 5 dummy variables. The one that you do not add to the model serves as reference. That can be any dummy variable.
could you show how to do multicollinearity test for the same data type?
The regression procedures for categorical dependent variables do not have collinearity diagnostics. However, you can use the linear Regression procedure for this purpose. Collinearity statistics in regression concern the relationships among the predictors, ignoring the dependent variable. So, you can run REGRESSION with the same list of predictors and dependent variable as you wish to use in LOGISTIC REGRESSION (for example) and request the collinearity diagnostics.
Hi Fawad, just got your nice message. Would you be able to tell me whether or not I have to convert my likert scale IVs to present and absolute values before running the ordinal logistic analysis (I read that somewhere). thanks
Hi. No. IVs can be continuous or discrete. DV in the ordinal regression has to be in ordinal.
Thank you
Hi Fawad, huge thanks for this video! I got one question, if i run my analyses i get the warning: "there are 118(66,7%) cells (i.e. dependent variable levels by observed combinations of predictor variable values) with zero frequencies. I checked the assumptions and found no values that are not being used. Can you explain if this is a problem, what's causing the problem and mostly how i can fix it if is a problem... Thank you very much again!
Hi. Are there any missing values in your data?
@@researchwithfawad Thanks for your answer, there are no missing values in my data. I checked for those...
so is the p-value important also when interpreting the parameter estimates?
Yes, it shows if the results are significant or not.
Can I use continuous variables in Ordinal Logistic Regression? I tried using it, but it seems like continuous variables are converted to nominal variables when looking at SPSS results. It means that when using a continuous variable in the model, I obtain too many parameters corresponding to each different value in this continuous variable while I expected there to be only one parameter corresponding to that variable.
The DV has to be on ordinal scale. IVs can be continuous.
Thanks a lot! Yet one clarification...In your depenedent variable you have 5 star rating there. yet only four's output was only there. what about the 5th one. IF the 5th one won't come in output; pls explain why?, Thanks in advance! :)
Thanks for watching. May be no body choose the option.
Hi sir. I have a question. In statistics non parametric tests are the ones used for ordinal data, and I've never really heard of central limit theorem applied to ordinal data.
How can we theoretically justify the use of Wald test in ordinal regression, since it's a parametric test?
In statistics, nonparametric tests are commonly used when working with ordinal data.
Parametric tests, such as the Wald test, make assumptions about the underlying distribution of the data. The Wald test is a parametric test commonly used in regression analysis, including ordinal regression, which assumes that the dependent variable follows a particular distribution (e.g., a logistic distribution in the case of ordinal regression).
The theoretical justification for using the Wald test in ordinal regression lies in the asymptotic properties of the estimator. The Wald test relies on the central limit theorem, which states that under certain conditions, the distribution of the estimator becomes approximately normal as the sample size increases. In the case of ordinal regression, the Wald test is based on maximum likelihood estimation, which has been shown to have asymptotic normality properties.
@@researchwithfawad thank you very much!
Can someone please tell me what it means if the threshhold is insignificant but the locations are?
Can you please provide some details? Are your referring to the fits?
Hi Prof. How to explain if all estimate (variable X and Y) in negative?
This means a negative relationship provided the IV is continuous.
Are there any assumptions that need to fullfiled before applying orfinal logistic regression??
Please refer to the test of parallel lines in the video.
Hi! what does 0a means in parameter estimates. Thanks for answering
0 would mean no effect.
@@researchwithfawad You are such a big help for my research! May Allah bless you! by the way is the result okay if the all parameter estimates are non significant? Thank for answering
Dr fawad can we compute/group multiple items of DV with Mode(frequency) for ordinal logistic regression?
If you take mean than it will no longer be ordinal.
OK dear. If take Median than remains catagorical?
Hi Prof, is it possible to check interactions between independent variables and how to conduct it for ordinal logistic regression?
Yes, you can create interaction effect. You may standardize continuous variables and create interactions. Or for categorical no need to standardize.
@@researchwithfawad Thank you Prof for your feedback
can you do another one, that is very detailed and from scratch for beginners, i know nothing about the model?
Thanks. If you are looking for model development, you may watch the video in the attached playlist.
How to Find Research Gaps/New Topic for Research: ua-cam.com/play/PLb7vm6tsQ3KslXSw1BOb3Q0yJnksLyi2Z.html
Hi sir,
Is the data "interest" in ordinal or scale? What happens if the data interest is ordinal, because the dependent variable should be ordinal so it uses ordinal logit regression
The dependent variable needs to be on ordinal scale.
Sir,can we apply OLS regression in determine impact of demographics (Age ,Income,gender, educarion)on dependent variable financial management behavior (Likert scale)...do we need to convert likert scale
If there a multiple items in the DV, you will need to take its mean and if the predictor categorical variables have more than two categories you will need to create dummy variables and use linear regression.
@@researchwithfawad yes sir i measure it with five item likert scale ..can i apply ols after taking mean
Yes
My significance level under Parameter Estimates is a dot "." Does it mean anything?
What is your sample size?
@@researchwithfawad Actually the sample size is small (90 observations) but I changed IVs from factors to covariates and now it seems much more clear although most effects are not significant
My question is when and how do one decide to use Linear regression or ordinal logistics regression sir?
Thanks I await your answer
Thanks for watching. Use Ordinal when the DV is a categorical ordinal variable and for continuous DV, use Linear.
Sir! how do we interpret the Likert scale. Kindly answer pls. my deadline is near huhu. Thanks btw
Which Likert Scale?
If your DV is on Likert scale use Multiple Linear Regression?
@@researchwithfawad My DV is ordinal, a Likert scale (not motivated to highly motivated) how do i know if male (or for married) for example are moderately motivated or fairly motivated)?
And whats the difference between parameter estimates and log of odd. thanks
Its urgent huhu thanks for answering! God bless
I have a question sir, what if one has an Independent Variable with 5 categories, ie, strongly disagree, disagree, neutral, agree and strongly agree. assuming strongly agree is set as the baseline/reference category, how can you interpret the estimate value if; strongly disagree is -ve sign, disagree is -ve sign, neutral is +ve sign, and agree is -ve sign while baseline is redundant (0)？and in another case strongly disagree is -ve sign, disagree is -ve sign, neutral is +ve sign, and agree is +ve sign while baseline is redundant (0)?
A positive or negative sign with the estimate in the Parameter Estimates table shows the likelihood of falling into a particular category. + sign is associated with an increase likelihood of case falling into a higher category in the dependent variable. - sign is associated with an increase likelihood of case falling into a lower category in the dependent variable. For instance, a Positive Sign with and IV with 5 Categories would mean that with increase in the value of that variable, there is a higher chance of falling into a higher category on the DV.
@@researchwithfawad Thank you for the response sir, very much appreciated. Perhaps my question was not clear enough, what I was intending to ask is if the variable has different signs at the same time at different categories. For instance a construct “happiness” with five categories and some categories have - sign while some have + sign in the estimates. How would one interpret happiness? When it has some categories that are + and some that are - in the estimate at the same time.
great work but you never spoke about the threshold
Thanks for watching. Which threshold please
There was a threshold value including its estimates. I think it is just above to the coefficient estimates
@@researchwithfawad
Can we Consider Exp(B) as an Odd ratio?
For Odds Ratio, please watch
ua-cam.com/video/tlslGucrD_c/v-deo.html
When interpreting the results shouldn’t we consider the significance value?
Yes we should
Help me to download the data file.
Please visit
researchwithfawad.com/index.php/download-research-resources/
Hi professor, in this video the DV is ordinal, but running the test in SPSS you have taken DV as ratio data instead of ordinal data, is this correct pls clarify, thank you in advance
Thanks for watching. In Variable view that is only for identification purposes and doesn't affect the analysis.
@@researchwithfawad thank you so much sir one request sir can you share any article with ordinal logistic regression for interpretation purpose pls
Please put in the following string in Google Scholar, and you will find a number of studies.
allintext: "Ordinal Logistic Regression"
@@researchwithfawad thank you professor
in goodness of fit, the significance for Pearson is coming 0.000 in my case and Deviance is 0.998. Is it acceptable?
No it is not acceptable. The difference should be insignificant.
Dr. I want your email if possible.
My email is kh.fawad83@gmail.com
@@researchwithfawad thanks dr
Just check your email Dr.