Names are easy. Use names that are familiar to your participants.

Faces may be harder, depending on what you are hoping to measure. If you are interested in face perception (a subset of visuospatial cognition based largely in the fusiform gyrus), then you want a series of face discrimination tasks of increasing difficulty. In other words, you begin each "item" with a target face, then present several faces (target plus 2-3 foils) and ask the participant to identify the target. At first the faces are clearly distinct, but they gradually get more similar until it reaches a point where very few people can make the distinction at better than chance levels. (Note that since this is a multiple-choice task, you need several items at each level of difficulty.) You need to ensure that there are no "obvious" cues that give away the correct answer: things like different skin tones, hairstyles, or articles of clothing/jewelry.

You could use Face Generation software (such as https://facegen.com/) for this purpose. These packages allow you to manipulate the degree to which faces are "related." So you can generate faces of people who have only random features in common, others who bear a cousin-like resemblance, others who are equivalent to full siblings, etc. You can also control skin tone.

You will be able to employ several conditions: recognition, immediate recall, and delayed recall.

I would advise against placing importance on names, by the way. If you ask participants to "remember" that this face bears that name, then you are also testing verbal memory.

All the programming languages use English letters, so you may use English names to accommodate the original programming language.

Great question

For names point, I think names should be in Arabic. If you use images, try to use 5 images for each face with some different features like angle of image capturing.

Thank you all for your concern and useful advice

Dear Stephen Joy,

I have tested spacial memory in mice. Now, I would like to test memory in human. I read that "face–name association memory is mediated principally in the hippocampal subregions, the cornu ammonis (CA) fields 2/3 and the dentate gyrus". Therefore, I am planning to use simple memory test that I can interpret its data as I have no experience in this field yet. Is my choice for face recognition test correct? or I have to choose another test, knowing that I would like to test either social or spacial memory in human that associated with the hippocumps.

Thanks in advance,

I do not pretend to be an expert in this specific research area. I am sure, though, that face-name association is not the same thing as face recognition or face memory. Face recognition would be a necessary first step. Therefore, I think an adequate research program would begin with face recognition, then proceed to face memory, and finally to association of faces with names. The last task isn't really "spatial" as such, but rather a form of associational memory with both spatial and verbal components.

If one only measured face-name association, one would not know the source(s) of any deficits. You see the problem? A person performs poorly. Is it because of deficits in associational memory? Memory for faces? Or basic facial recognition? Think of it as serial processing.

(To be complete, one would also want to assess purely verbal paired-associates learning as well. Again, this would be a kind of "baseline" for the higher-order task.)

If, however, you only want to study the association of names with faces, then I recommend making the faces you use as distinct as possible. Do not make this part of the task the least bit difficult. Ignore everything I said before about including faces that are increasingly similar, eliminating skin and hair differences, etc.

Thank you for your comprehensive answer.Your answer alerted me to many points that I had not noticed before.

Thanks million