How can a researcher increase reliability? Consider what you might do if you find the 'right' test for your dissertation, but cannot find enough evidence to support strong reliability. What can you do in the planning stages to improve the reliability of the test and/or improve the overall outcome of your project?