If only one item is identified for a mediating variable, is it an acceptable solution? In factor loading it is getting a reasonably high value (> 0.8). Are there references of single item variables being used in the past? In my research am facing this scenario - looking for a resolution.

Thanks