Design list test
From SiteRay wiki
| SiteRay Test | |
|---|---|
| First appeared: | Sitescore 3.0 |
| Applies to: | All versions |
| Type: | Individual test |
| Scored: | Never |
What does it do?
Identifies separate graphic designs / layouts used by a website, and identifies how frequently each is used.
Why is it important?
This test is useful for several reasons:
- In large sites, it is common to have 'left over' pages from a previous design that were missed during an upgrade. This test will find these.
- Unexpected error messages, or broken pages, often appear as different designs (especially where the design is broken). This test identifies those pages.
- When considering the replacement or upgrade of a website, this test can give an at-a-glance overview of the site is comprised, without having to view every page individually.
Example results
How is it measured?
Sitescore analyses the following aspects of each page:
- Images used
- CSS used
- HTML / XHTML source, traits of
- Framesets, whether they are used
These are combined and weighted. Sitescore then sorts these into sets, each of which represents a design.
The thresholds and precise algorithms used for this test are trade secrets. We revise this algorithm from time-to-time to better identify similar and different page designs.
Users may specify their own overall threshold level for design similarity, if they would prefer to see more distinct or similar designs listed.
Common problems
This test is an automated summary of a subjective trait: graphic design. Human beings would be unable to agree, in many instances, how many 'designs' a website has and what they are. Therefore a computer - without the benefit of human intelligence - has a particularly hard task. Our approximation is however very accurate for the vast majority of cases.
It is possible for Sitescore to report results that you may disagree with. In nearly all cases, changing the Design similarity threshold setting will correct these problems.
How to use this test effectively
This test can be reviewed on an infrequent basis, particularly prior to or just after a redesign. Results should be investigated to ensure that any designs found are authorised and appropriate.
