I have a list of 40,000 words that I'd like to test in a survey. But every respondent should only be asked about 20 of those. So it would look something like this:

Respondent 1: “is w1 a real word?”; “is w2 a real word?”,…, “is w20 a real word?”.

Respondent 2: is w21 a real word?” etc.

Word subsets do not have to be drawn randomly, they can be presented in order (1-20, 21-40, 41-60, etc.).

I have read through the answers in:

However, I'm wondering if there is an easier alternative, 4 years later. I have tried the loop and merge function, but all it does is sample with replacement from the list of 40,000 words, so I get some words that are shown multiple times to various respondents and some words not shown at all. I am unfortunately not a programmer. Thank you so much.