I usually never comment on UA-cam, but this video solved my problem that I didn't even know how to Google correctly to find the answer. Many thanks! You're saving our academic degrees :D!
Yes, you can use syntax to combine the columns. This method was covered on this research gate thread: www.researchgate.net/post/How_do_I_combine_2_variables_in_SPSS_into_one_and_copy_the_values_from_the_original_variables
Hi there, is it the same when working with nominal/categorical data? I've done a survey through Qualtrics for psych honours project using all nominal data. When importing from Qualtrics into SPSS participants responses for every question and condition are coming up with their own column (Multiple columns for each Question, and responses are either 1 or a dot). How do I combine categorical responses into one column, so that I can then assign values? At the moment each categorical response all have the value of "1" so I'm unsure how I would be able to tell what category they were originally in if I use a "sum" option.
Hi there! This trick is for numerical data. It won't work for nominal data, but there are other tricks you can do. Perhaps the simplest if you have few participants is sorting the data and then a quick copy and paste, if I am honest with you :) Feel free to send me an email of you have additional questions (meg.huntoon@gmail.com).
Hi Meg! Thank you so much for this helpful video. I still have a problem merging the columns: so I have 3 conditions that I 'd like to merge. The problem is that once they are combined, two of the conditions are identified by number 1, while the other one is 2, so I can't label them correctly. Before merging, all these conditions are shown by 1 (basically once the participant chose one of these three responses, it is shown as 1 in each of these columns. Could you let me know how we recognize these answers and label them once merged together?
Hello! Thanks for your comment! I would change one of them to a different number before the merge by recoding them. That should do the trick :) email me if you have questions about how to do that (meg.huntoon@gmail.com)!
Hi thank you for the helpful video. However, I have difficulties to combine two variables. I have two questions in my survey regarding the educational level (1. Which education are you currently following?, Which is your highest level of education achieved?) Both questions are categorized in level low, middle, high. Now I would like to combine these questions. For example 5 participants low (first question), 3 participants low (second question) --> 8 participants low. I am looking forward to your response. THNANK YOU! Johanna
Hi! Thanks for your comment! If I’m understanding correctly, you’re looking to create a count of the number of „low“responses? If so, check out this video: ua-cam.com/video/YQSKLLlJSeg/v-deo.html If that’s not quite what you were looking for, send me an email and I’ll try to help! Meg.huntoon@gmail.com
Thank you this was SO helpful! I'm just wondering if there are any other ways to do this? For example, I had dataset with the variable "height" measured randomly in cm or ft (and another with "what unit did you choose") and needed to create a new column with everything in cm. I could not find a way to do multiple IF statements without using Syntax (I'm extremely new to SPSS and stats software in general) and ended up just using compute variable twice with different functions to make 2 columns with measurements in cm then merge them like this. It worked perfectly but maybe not the most efficient...
Thank you for watching! It depends on how your variables are set up (are they string variables or numeric) and is there data in both variables that you are hoping to combine. If you would like, send specifics about your data to my email (meg.huntoon@gmail.com) and I can help you decide which method will be best for your project.
And I figured it out... you need to select "String" for output type with you select compute variable, make sure the length fits the length of your string then use CONCAT(V1, V2, ...)
Great video! Unfortunately when I do it this way I get the comment: “The numeric argument required for the function specified was not supplied.” What am I doing wrong?
Hi Ruben! Thanks for watching. Can you post or e-mail (meg.huntoon@gmail.com) me the formula you typed into the compute variable box? Is it possible that you are missing a parenthesis or comma or have an equal sign that isn't needed?
Hi Megatron 24, I have a quick question. I have a 2 by 2 + 1 experimental design. I created two independent variables. However, I don't know how to code them. The levels of my first IV are either whether a message is serious or humorous. The levels of my second IV are either whether a solution is offered in the message or not. Moreover, one of the groups is an isolated control group that was not exposed to these treatments. What values do I put for this group?
Naomi van Bree Hello! Thank you for your comment. Are you trying to combine the separate columns in your data like is shown in the video? If so, then this method will work the same as shown with your design, only you will need to add all the columns for each DV (in your case, three for each). If instead you are hoping to code your IV to fit with the analyses you want to run, then how you code them will depend on what you are hoping to do with the data. Does that help?
@@huntoon6393 Thank you for your quick reply! Firstly, I created a "condition" column like you showed in one of your previous videos. I have 5 conditions in total, one of which is the isolated control group that was not exposed to the treatments. Then I created two new variables for my IVs. However, I don't know how to code these variables. Should I use effect coding like 1 for humor and -1 for serious or dummy coding like 1 for humor and 0 for serious? Or can I do it like you did it in your previous video (1 = humor, 2 = serious). I want to perform a 2 by 2 ANOVA in which I compare the means of my treatment groups (and the interaction between my IVs). But I also need to compare my control group with my treatment groups. However, basically the data for my control group is "missing" for my IVs, since this group was not exposed to any of the treatments. I don't really know what value to put for this group in my two columns for my IVs. The same goes for one of my dependent variables. I only measured it for my treatment groups but not for my control group. If this isn't clear, is there a way I can maybe e-mail you to show some screenshots and explain more clearly what I mean?
Naomi van Bree Naomi van Bree That makes sense. If you have a priori hypotheses, I would use planned comparisons to analyze your data. I don’t have a video on that type of analyses, but there are many of them out there. If you want to run a regular anova, I would code the control condition as 0 in one of your IVs. In that case, you will have a 2(humor vs none) x 3 (solution vs none vs control). I believe running the analyses in this way will provide you the tests you are hoping for. One thing I am struggling to understand is what your control condition adds beyond the group who received no humor and no solution, but I am likely just not picturing the design correctly in my head. Was it that the control condition didn’t receive any stimuli?
@@huntoon6393 My research is about webcare responses. Basically, one group received a message with humor but without a solution, another group received a message with humour and with a solution, another group received a message without humor (serious) without a solution and finally one group received a serious message with a solution. The control group did not receive any message. This group was therefore not exposed to the two IVs. I created the two IVs in my dataset. Since my control group wasn't exposed to any of the IVs at all, I'm not sure what value I need to put for this group in my two IV columns. If, for example, I code humour versus serious with 1 and 0 (e.g. 1 = humour, 0 = serious), how do I code the group that did not get exposed to humour at all? What value do I put in there?
Naomi van Bree Ok. That makes sense. You can code the control group as I suggested in my previous reply to run an anova. The output will give you comparisons and interactions for each group. You will also want to request the means so that you can tell the direction of effects. And of course follow up any significant interactions with post hoc tests.
in the middle of doing my MSc dissertation and this totally saved my life
I usually never comment on UA-cam, but this video solved my problem that I didn't even know how to Google correctly to find the answer. Many thanks! You're saving our academic degrees :D!
I waisted an afternoon trying to get this to work before watching your video, you just went to the top of my Christmas card list.
Omg, I just spent 2hrs trying to figure this out and you saved me!!! Thank you. 😅❤
Most helpful ever! saved my master thesis literally.. Thank you so much !
Most helpful vid ever!!!! Haha exactly what I needed for my thesis. Thank you so much!
Helpful. Is there a way to just merge the columns without sum?
Yes, you can use syntax to combine the columns. This method was covered on this research gate thread: www.researchgate.net/post/How_do_I_combine_2_variables_in_SPSS_into_one_and_copy_the_values_from_the_original_variables
Soo helpful. Thanks Meg
Hi there, is it the same when working with nominal/categorical data?
I've done a survey through Qualtrics for psych honours project using all nominal data. When importing from Qualtrics into SPSS participants responses for every question and condition are coming up with their own column (Multiple columns for each Question, and responses are either 1 or a dot). How do I combine categorical responses into one column, so that I can then assign values? At the moment each categorical response all have the value of "1" so I'm unsure how I would be able to tell what category they were originally in if I use a "sum" option.
Hi there! This trick is for numerical data. It won't work for nominal data, but there are other tricks you can do. Perhaps the simplest if you have few participants is sorting the data and then a quick copy and paste, if I am honest with you :) Feel free to send me an email of you have additional questions (meg.huntoon@gmail.com).
Great video!! Thank you!
Wow this saved my life!!! Thank you so much!!!
Hi Meg! Thank you so much for this helpful video. I still have a problem merging the columns: so I have 3 conditions that I 'd like to merge. The problem is that once they are combined, two of the conditions are identified by number 1, while the other one is 2, so I can't label them correctly. Before merging, all these conditions are shown by 1 (basically once the participant chose one of these three responses, it is shown as 1 in each of these columns. Could you let me know how we recognize these answers and label them once merged together?
Hello! Thanks for your comment! I would change one of them to a different number before the merge by recoding them. That should do the trick :) email me if you have questions about how to do that (meg.huntoon@gmail.com)!
Hi thank you for the helpful video. However, I have difficulties to combine two variables. I have two questions in my survey regarding the educational level (1. Which education are you currently following?, Which is your highest level of education achieved?) Both questions are categorized in level low, middle, high. Now I would like to combine these questions. For example 5 participants low (first question), 3 participants low (second question) --> 8 participants low. I am looking forward to your response.
THNANK YOU!
Johanna
Hi! Thanks for your comment! If I’m understanding correctly, you’re looking to create a count of the number of „low“responses? If so, check out this video: ua-cam.com/video/YQSKLLlJSeg/v-deo.html
If that’s not quite what you were looking for, send me an email and I’ll try to help! Meg.huntoon@gmail.com
Thank you this was SO helpful! I'm just wondering if there are any other ways to do this? For example, I had dataset with the variable "height" measured randomly in cm or ft (and another with "what unit did you choose") and needed to create a new column with everything in cm. I could not find a way to do multiple IF statements without using Syntax (I'm extremely new to SPSS and stats software in general) and ended up just using compute variable twice with different functions to make 2 columns with measurements in cm then merge them like this. It worked perfectly but maybe not the most efficient...
I wonder if I want to combine two variabel into one, ex. Var A: 1 var: 2 & I want the new variabel value is 12 not 3. How do you do that ?
Thank you for watching! It depends on how your variables are set up (are they string variables or numeric) and is there data in both variables that you are hoping to combine. If you would like, send specifics about your data to my email (meg.huntoon@gmail.com) and I can help you decide which method will be best for your project.
Genial!!! No sabía como realizar la agrupación. Muchas gracias por subir tu video.
Is there a similar function to deal with "String" data? Or does this only work for "Numeric" Data?
And I figured it out... you need to select "String" for output type with you select compute variable, make sure the length fits the length of your string then use CONCAT(V1, V2, ...)
The Funky Deep great! Glad you figured it out! Good luck with your work.
@@huntoon6393 thanks!
Thank you.
Great video! Unfortunately when I do it this way I get the comment: “The numeric argument required for the function specified was not supplied.”
What am I doing wrong?
Hi Ruben! Thanks for watching. Can you post or e-mail (meg.huntoon@gmail.com) me the formula you typed into the compute variable box? Is it possible that you are missing a parenthesis or comma or have an equal sign that isn't needed?
Already solved the problem, still thank you for your response!
Great!! Glad you fixed it!!
Hi Ruben, I ran into the same problem. Could you please share how you resolved it? Thanks!
Hi Megatron 24, I have a quick question. I have a 2 by 2 + 1 experimental design. I created two independent variables. However, I don't know how to code them. The levels of my first IV are either whether a message is serious or humorous. The levels of my second IV are either whether a solution is offered in the message or not. Moreover, one of the groups is an isolated control group that was not exposed to these treatments. What values do I put for this group?
Naomi van Bree Hello! Thank you for your comment. Are you trying to combine the separate columns in your data like is shown in the video? If so, then this method will work the same as shown with your design, only you will need to add all the columns for each DV (in your case, three for each). If instead you are hoping to code your IV to fit with the analyses you want to run, then how you code them will depend on what you are hoping to do with the data. Does that help?
@@huntoon6393 Thank you for your quick reply! Firstly, I created a "condition" column like you showed in one of your previous videos. I have 5 conditions in total, one of which is the isolated control group that was not exposed to the treatments. Then I created two new variables for my IVs. However, I don't know how to code these variables. Should I use effect coding like 1 for humor and -1 for serious or dummy coding like 1 for humor and 0 for serious? Or can I do it like you did it in your previous video (1 = humor, 2 = serious). I want to perform a 2 by 2 ANOVA in which I compare the means of my treatment groups (and the interaction between my IVs). But I also need to compare my control group with my treatment groups. However, basically the data for my control group is "missing" for my IVs, since this group was not exposed to any of the treatments. I don't really know what value to put for this group in my two columns for my IVs. The same goes for one of my dependent variables. I only measured it for my treatment groups but not for my control group. If this isn't clear, is there a way I can maybe e-mail you to show some screenshots and explain more clearly what I mean?
Naomi van Bree Naomi van Bree That makes sense. If you have a priori hypotheses, I would use planned comparisons to analyze your data. I don’t have a video on that type of analyses, but there are many of them out there.
If you want to run a regular anova, I would code the control condition as 0 in one of your IVs. In that case, you will have a 2(humor vs none) x 3 (solution vs none vs control). I believe running the analyses in this way will provide you the tests you are hoping for.
One thing I am struggling to understand is what your control condition adds beyond the group who received no humor and no solution, but I am likely just not picturing the design correctly in my head. Was it that the control condition didn’t receive any stimuli?
@@huntoon6393 My research is about webcare responses. Basically, one group received a message with humor but without a solution, another group received a message with humour and with a solution, another group received a message without humor (serious) without a solution and finally one group received a serious message with a solution. The control group did not receive any message. This group was therefore not exposed to the two IVs. I created the two IVs in my dataset. Since my control group wasn't exposed to any of the IVs at all, I'm not sure what value I need to put for this group in my two IV columns. If, for example, I code humour versus serious with 1 and 0 (e.g. 1 = humour, 0 = serious), how do I code the group that did not get exposed to humour at all? What value do I put in there?
Naomi van Bree Ok. That makes sense. You can code the control group as I suggested in my previous reply to run an anova. The output will give you comparisons and interactions for each group. You will also want to request the means so that you can tell the direction of effects. And of course follow up any significant interactions with post hoc tests.
This was helpful! thank you!
Very helpful. Thanks you
Thanks, it was very helpful
Thank you!
thankx!
Hi have sent you an email :)