Scientists randomly select ten groups from a population of men over 50 years old. They calculate the mean weights of each of these groups. The variability between these means can be best attributed to

Notice that the subjects of the study are men over 50 years.

So, you need to consider that each subject is a different person with different weight and other different characteristics that determine that weight.

If we have 10 groups randomly selected, that means there's gonna be variability, otherwise, the study wouldn't make sense.

So, if each person is different, then their weights is gonna be different, giving different means.

It would be rare if both means result equal, beacuse that would mean the subjects had the same weight.

Therefore, the variability between these means can be best attributed to the variability of subjects, the difference between each person.

From a numerical perspective, the variability is attributed to the spread of the data group, that is, if there is too much difference between weights, the variabilty will increase.