I do not think SPSS can create these on input (except in the New Data Enrtry system of SPSS, but I do not think you will have this).

You can make a syntax with compute statements that can do this. It runs like

IF (Char.Index(string1, "yellow")> 0) ColorNr = 1.

You have to do this for each color and each variable so like

IF (Char.Index(string1, "yellow")> 0) ColorNr = 1.

IF (Char.Index(string1, "green")> 0) ColorNr = 2.

IF (Char.Index(string2, "yellow")> 0) ColorNr2 = 1.

(I hope I have used the correct syntax as I do this on holiday without my syntaxes to look it it up). If you have problems, let me know and I will look it up. But start with the menu and paste it to a syntax.

Hello Peter,

Thank you for your answer.

i actually did that. And it works, but only for one variable... so the broblem is that when I have the case: string1: yellow+brown+blue , my wanted out put should be: 1, 2, 3 and instead I only have 1 as output.

Do you know how can I solve this issue?

I do have a solution, but maybe you want something else. I do not think you want to have a new string that makes 'yellow+brown+blue' '1,2,3'.

Maybe you might want:

IF (Char.Index(string1, "yellow")> 0) ColorYellow = 1.

..... other color codes

IF (Char.Index(string1, "green")> 0) ColorGreen = 1.

IF (Char.Index(string2, "yellow")> 0) ColorYellow = 1.

..... other color codes

IF (Char.Index(string2, "green")> 0) ColorGreen = 1.

IF (Char.Index(string3, "yellow")> 0) ColorYellow = 1.

..... other color codes

IF (Char.Index(string3, "green")> 0) ColorGreen = 1.

RECODE ColorYellow To ColorGreen (SYSMIS = 0) .

I assume that stringn are in lower case, otherwise the syntax becomes a little more complicate: IF (Char.Index(LOWER(string1), "yellow")> 0) ColorYellow = 1.

You could also use a few statements to make lower cases of the string1 to string3.

COMPUTE string1 = LOWER(String1) .

COMPUTE string2 = LOWER(String2) .

COMPUTE string3 = LOWER(String3) .

It can be even more compact if you use DO REPEAT .

COMPUTE string1 = LOWER(String1) .

COMPUTE string2 = LOWER(String2) .

COMPUTE string3 = LOWER(String3) .

DO Repeat QQ = string1 string2 sting3 .

IF (Char.Index(QQ, "yellow")> 0) ColorYellow = 1.

..... other color codes

IF (Char.Index(QQ, "green")> 0) ColorGreen = 1.

END REPEAT.

RECODE ColorYellow To ColorGreen (SYSMIS = 0) .

EXECUTE .

The 'do repeat'-statement does all the transaction for all variables at ones, but it need to finish with a 'END REPEAT'-command statement.

You will have as many variables as you have defined colors. You can use MD-tables to analyze this kind of data.

Thank you but unfortunately i already managed to create my 'dummy variables' with 1 or 0 values... but i cannot analyze them because I only have dicotomic values... That' s the reason why I would need to create a system to assigne different number to different variables.. In the end, i'm gonna need to create a chart that shows which colors were there (with their frequencies) in the different strings.. also, how can I analyze the frequencies of the values with this type of dicotomic values?

Do you have the SPSS module 'tables'? You can create MD-variables for 0-1 codings or MR-variables for categorical values (like color1 = 1 (red), color2 = 1 (red) and color3 =1 (red). You can tell CTABLES to use cases or answers as reference categories. With multiple answers in multiple variables you can not have one variable to hold all information at ones. That is why MD-variables were created in SPSS. In CTables you can have several statistical tests. And your MD-tables can be turned into a charts showing how many times the colors were named. Does it have meaning that string1 is red and sting2 is red as well? You could have multiple MD-variables, but the matrix behind it would be very sparse if you have many colors.

i don't think I have this function, if I click on tables my spss only gives me: custom tables or multiple response sets... I'm not finding 'module tables'.

If you have 'custom tables' and 'mutiple response sets', the module tables is installed. It is a module that does not belong to 'standard' SPSS. But many universities may have this add-on installed. With multiple response sets you can create MD or MR variables and use them to create cross-tabulations. Be warned: if a case has not selected any option, it will not show in a MD or MR variable. You have to create a 'missing' category. A variable or value indicating 'none of the above' (after the movie 'Brewsters Millions'). If nothing has been selected you could have a variable ColorNone = 1 or have ColorNr =99 (where 99 stands for a 'blank'-color).

A MD syntax will look like:

*Define Multiple Response Sets.

MRSETS

/MDGROUP NAME=$concept1 LABEL='label concept 1' CATEGORYLABELS=VARLABELS VARIABLES= var01 to var 99

VALUE=1

/DISPLAY NAME=[$concept1].

After this command, you will have this variable added to your variable list in/for custom tables, but not in the variable-list for FREQUENCIES.

Dear Giorgia, have you checked the answer for your question in the other thread? Does it work?

https://www.researchgate.net/post/How_to_convert_String_into_Numeric_form_in_SPSS_without_losing_the_data