Evaluation of Websites.html (3426B)
1 <?xml version="1.0" encoding="UTF-8"?> 2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 3 <html><head><link rel="stylesheet" type="text/css" href="sitewide.css"><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/><meta name="exporter-version" content="Evernote Mac 6.13.3 (455969)"/><meta name="altitude" content="-0.5099493265151978"/><meta name="author" content="Alex Balgavy"/><meta name="created" content="2018-01-24 1:00:21 PM +0000"/><meta name="latitude" content="52.37356293843497"/><meta name="longitude" content="4.836268901730849"/><meta name="source" content="desktop.mac"/><meta name="updated" content="2018-01-24 1:17:32 PM +0000"/><title>Evaluation of Websites</title></head><body><div><b>Why evaluate?</b></div><div><ul><li>attract more visitors</li><li>sell more products</li><li>decide which web app to use</li><li>improve visitor ratings</li><li>etc.</li></ul><div><br/></div></div><div><b>Evaluation studies</b></div><div><ul><li>always start with a clear research question</li><ul><li>aka “problem statement”</li><li>practical/theoretical relevance, feasible</li><li>types of studies</li><ul><li>explorative — what is related?</li><ul><li>e.g. why do people visit the site again?</li></ul><li>descriptive — what happens?</li><ul><li>e.g. how many people find the site through a SE?</li></ul><li>explanatory — why does it happen?</li><ul><li>if you add login, do people visit the site again?</li><li>does a change in structure make it easier for visitors to find what they’re looking for?</li></ul></ul></ul><li>make a hypothesis</li><ul><li>a prediction of outcome of test</li><li>deduced from theory or observations</li></ul><li>collect data</li><ul><li>e.g. in lab experiments, survey, interview…</li><li>qualitative (non-numerical) and quantitative (numerical)</li><li>test dependent var with respect to independent var</li><li>specific population (customers/all web users/registered users/whatever)</li><li>specific sample (random/convenience/volunteers)</li></ul><li>evaluation methods</li><ul><li>common</li><ul><li>mockups</li><ul><li>low fidelity — early in design phase, only basic functionality, static, cheap. focus on concepts.</li><li>high fidelity — later in design phase, refined details, expensive.</li></ul><li>prototypes — working example of website</li><li>focus groups — moderated group discussion, early in design stage.</li><li>card sorting — group of people sort items into clusters to get intuitive structure for website</li><li>usability inspection — go systematically through website, check against Ten Web Guidelines. performed by dev team.</li><li>group walkthrough — group of people walk through website as if performing primary tasks</li><li>user testing — remote, observe user while primary tasks are performed. log actions, eye tracking, record video/audio.</li><li>survey</li><ul/></ul><li>specific to web evaluation</li><ul><li>web analytics</li><ul><li>analyse logfiles</li><li>JavaScript page tagging to capture visitor data</li><li>very good objective data, but privacy concerns and no insight into motivation or unvisited pages.</li></ul><li>online experiments<br/></li><ul><li>distribute visitors over versions, see which performs better</li><li>after release</li><li>example: A/B testing</li></ul></ul></ul></ul></div><div><br/></div></body></html>