It doesn't really matter what value you use for the constant in the Scherrer equation as long as you state it in any publication and are aware of the fact that strain and instrumental broadening of the peak(s) is not assessed in the Scherrer equation. Do not use different values for materials in the same family and you'll be able to compare. Most people will use a value around 0.9 although (in theory) the variation could be high.
You may be better utilizing Williamason-Hall.
There's a bunch of discussion on RG on this exact topic. Start here: