first you need to know what filetype the signals sound files are (guessing mp3)
for archive purposes save recordings as wav file regardless. Since you can change the file later, but you need the wav so it saves as the highest possible quality.
recording? You need to either set up a mini recording booth to remove noise and sound bounce (some ppl record in a padded cardboard box).
For mics use a condenser mic and run a sound test on input sensitivity to what your kids would say the line without clipping the audio. This is done by setting it to its highest and working your way down (always set the actual recording a slight bit less than what the balance is to compensate for sudden fluctuations in volume upon delivery).
e.g. its reccomended to be set high (to get best quality and clearest sound) but a notch lower than what you think. clipped audio is useless audio.
As for the actor? You are gonna have to teach your kids to talk into the mic a lil away from the mic and then project the voice as to envelope the mic with sound. But not only that but to avoiding delivering the line with breath (some breathiness and pops caused by bursts of breathing hitting the mic can be avoided with a pop filter, but its always best to merely practice good technique in the first place)
After you are happy with your audio recording. Do a bit of cleaning up noise (ideally you avoid it by doing a proper sound proof set up).
normalise the audio before exporting.
quick sound test.
voila. You got basically studio quality voice recordings.
