This is for gain. It doesn't set the signal to that value > 0.0 dB, it only amplifies the signal if > 0.0 dB. If some input signal is very low, for example at -30 to -40, and your other sources are normal, in the -20 area, it makes no sense to reduce all sources to -40 to make them equal, but to amplify the -40 signal to -30 or -20. This is still below 0 dB, so no clipping will occur. Of course, if you amplify a signal that is already loud, it can go over 0 dB and get clipped.
If you have a very quiet signal, try to not use this amplification in the first place. Instead, look into the device that generates the signal. Try to make the device send a higher volume. If it is a windows audio device, you can look into the second tab ("Levels") of the old control panel configuration. Make sure the level is high - ideally near 100. These settings can be seen as attenuator: 100 is 100% of the signal, thus the full signal. 50 is 50%, thus the signal reduced to half.
Try to avoid you reduce (attenuate) some signal near its creation and re-amplify it later during processing.
Distortion and clipping appear the same to the ear but have different causes. Distortion is mainly from analog audio processing. The frequency response is not linear with analog audio processing. The analog components have a optimal work area, where the frequency response is almost linear. Very loud signals are outside the work area. Outside the work area, a signal becomes distorted due to the inadequacy of analog processing. If you first amplify an analog signal so it gets distorted, and later attenuate so it is quiet enough, the signal is still distorted.
Very low signals should also be avoided. You always have a small bit of noise with very low signals. If you amplify low signals, you amplify the noise as well, so it becomes audible.
The best analog work area is just below the green/yellow border. For digital signal processing, the frequency response is linear over the whole volume area from -infinity to 0 dB.