Please give proper explanation and Typed answer only.

Susan is conducting marketing research for an apparel company in Texas, which is interested in determining attitudes towards a planned new logo. In Amazon Mechanical Turk, respondents saw the logo and rated it on a variety of measures, which were averaged to form a score. Despite pretesting the same survey four times on the same state, susan finds that the average is not converging, with each sample yielding completely different average scores. This implies that Susan logo attitude scale has: Low reliability or Low validity