First, you SHOULD not cal a refract with RO/DI water. Hand-Held refracts, such as this one, are set to "salt water" and not seawater. Meaning, water that has table salt in it. The RI (Refractive Index) of seawater is NOT the same as the RI of salt water. On a Lab Grade refract, which these cheap ones are not, they will always read .0015 to low. That means if you are testing seawater, with that refract calibrated with RO/DI, it needs to read 1.028 to equal 1.0265, as 1.0265 + .0015 = 1.028. The issue with these cheapo's is the quality of the optics. We have found them to be off from a NSW std 1.0265 from ~1.023 ( 31 ppt) to 1.0265 (35 ppt). That is allot, as much as ~ 4 ppt or 0.003 off and not +/- .001 or Sg or +/- 1 ppt as Salinity. If you cal them in Pinpoint 53 mS. a mock seawater, you can get them right on the nose @ 1.0265. They will be fairly accurate for a range of about 1.023 - 1.030.
Your 1.024 vs 1.025 that falls within the accuracy range for either of +/- 0.001. When IO hydrometers are not properly cared for they will read low, which is what you may be seeing here. If one lets seawater dry out in it that salt collects on the needle making it heaver. Thus, it reads lower than it should. When ever these swing arms types are used, they should be washed well with RO/DI water, left wet/damp and put in a plastic zip lock bag. And then soaked in a pH 4 Buffer, like those for cal pH probes, ovenight 2-4 times / year.
edit:
I'm having a bad day on the key board and with the head. So, this has been edited about 5 flipp'in times