upa - home page JUS - Journal of usability studies
An international peer-reviewed journal

User Experience and Accessibility: An Analysis of County Web Portals

Norman E. Youngblood and Susan A. Youngblood

Journal of Usability Studies, Volume 9, Issue 1, November 2013, pp. 25 - 41

Article Contents


Results

We identified 39 county web portals. As we describe in the Methods section, one was not functional during the study. Over half of the Alabama counties, 58.2%, have adopted web portals for e-government. The results support portal adoption hypothesis: Web portal adoption is weakly correlated with population and moderately correlated with income (see Table 1).

Table 1. Correlation Between Web Portal Adoption and Population and Income

Table 1

The usability hypothesis asked how well Alabama county websites adhere to basic and broadly accepted usability standards. Sites adhered to some of the standards and best practices, but many were lacking in multiple ways. Table 2 provides detailed results of the measure-by-measure adherence to usability practices.

Table 2. Usability of County Sites

Table 2

* Always assessed on internal pages.

The findings in this portion of study were at least partly in line with those in the Youngblood and Mackiewicz (2012) analysis of Alabama municipal websites. By-and-large, the county websites included a link labeled home at either the top of left-hand navigation or at the left of the top navigation (81.2%). Sites without the labeled link either used a house-shaped icon or relied on the user knowing to click on the logo or header graphic. With only three exceptions, the county sites placed their navigation in the recommended locations—the left and/or top of the page. County sites also typically fit within the suggested page width of 1024 pixels (94.7%). The main county pages were also easy to locate with the Google search results—all appeared within the first 10 (first page) of search results. Most sites (81.6%) used text links that were a different color than the surrounding text. Substantially fewer had links that changed color after being visited (39.5%) or were underlined (52.6%). On average, sites met between 10 and 11 of the 14 tested usability criteria.

We expected the results to support the null hypothesis that usability would not correlate with demographic factors, and indeed neither the usability indexes (the composite usability scores) nor the individual usability variables correlated with the demographic variables. For two-tailed tests (df=36), r and rpb were less than the critical value of 0.320.

We then addressed adherence to best practice coding, including accessibility standards, valid HTML, and the use of external style sheets (see Table 3). The majority of Alabama county sites did not have valid code; only 10.8% had no validation errors. The same was true for accessibility errors when tested in WAVE; only 18.4% were error-free when we included the two mostly Flash-based sites in the analysis (n=38). Even excluding those sites (n=36), the passage rate is still low: 19.4%. For detailed accessibility results, see Table 4. However, the majority (76.3%) did have external style sheets.

Table 3. Overall Accessibility, Validation, and Style Sheet Use of County Sites

Table 3

* The W3C validation service was unable to run on one county site despite multiple attempts.

Table 4. Error-by-Error WAVE Performance of County Sites

Table 4 Table 4

The results supported the null hypotheses for part of the best-coding-practices hypothesis: There was no significant correlation between the demographic variables and either W3C valid code or the use of external style sheets. However, there was an unexpected negative correlation between passing WAVE’s accessibility check and per-capita income, which means that the lower the income, the higher the number of errors (see Table 5).

Table 5. Correlation Between Passing WAVE Evaluation and Per-Capita Income

Table 5

*significant

The last hypothesis asked whether the county websites were prepared to take advantage of new communication technologies such as mobile devices. For the most part, they were not, with only 15.8% being prepared (See Table 6).

Table 6. Mobile-Device Readiness of County Sites

Table 6

We expected no correlation between adoption of new technologies and either county population or income. As expected, media styles did not have significant correlations with demographics.

 

Previous | Next