Technology Information/Requirements

  • Client Platform: Web browser-based online assessment - should work with recent versions of Internet Explorer and Firefox (these browsers have been used during development work)
  • Add Ons: Adobe/Macromedia Flash 9 or better required
  • Minimum monitor resolution: 800 x 600 pixels, 1024 x 768 is preferred
  • JavaScript: must be allowed in browsers used to access the test(s)
  • Cookies: Session state cookies must be allowed in browser used to access the test(s)
  • Cache: Online web based testing content must NOT be cached. Every effort must be taken to NOT cache test content on both network appliances and workstations. Systems must be monitored for cached content and purged immediately after testing if found.
  • Proxy Servers: Proxy servers cannot be set to cache the content of the test. The test is dynamic content targeted at specific users logged into the test. Proxy servers should be set to pass regular (port 80) and encrypted traffic (port 443) from everything on data.ncsu.edu.
  • Firewalls & Content Filters: Firewalls and content filters need to allow regular http (port 80) and encrypted https (port 443) from data.ncsu.edu
  • Pop-up Windows/Pop-up Blockers: Due to the current design of NCTest, pop-up ad blockers WILL CAUSE PROBLEMS. Pop-ups must be allowed from data.ncsu.edu in order to force the browser into full-screen mode. Pop-up windows must be allowed to open in a new window. Pop-up windows must NOT open in a tabbed window.
    • Pop-Up Policy Problem in Microsoft Internet Explorer 7: When taking Online Tests the test opens in a tab versus a new window without navigation bars, icons, etc. 
      Cause: IE7 has a software setting to force IE to handle pop-ups in a particular way.
      Solution: Go to Tools, Internet Options, Under General select Setting by Tabs, find When a pop-up is encountered and select Let Internet Explorer decide how pop-ups should open. Click Ok.
  • Test Prohibitions: Test takers WILL NOT have to move to windows outside of the test administration window.
  • Test Navigation: Soft section breaks - test takers can navigate forward and backward across the entire test until completed
  • Question Response Communications: Responses to test questions are sent back to a centrally hosted system at NC State University that resolves to data.ncsu.edu. Each time a student responds to a question and clicks a navigation button the response is transmitted to the server. If a student changes a response to a question previously answered and clicks a navigation button, then the change in the response is sent to the server.
  • Pre-Testing Checks: Before testing, please ensure that the local network operates without problems such as intermittent loss of connectivity, and that bandwidth intensive applications such as daily backups, network virus scans, etc. will NOT be scheduled to run during the test. Steps should be taken by the school/system to make technology staff and Internet service providers aware of the testing schedule, so that they are supplying sufficient bandwidth to the school/system to accommodate the test and that they do not schedule special events or maintenance during testing.
  • Host Service: The Technical Outreach for Public Schools (TOPS), based at North Carolina State University ensures that adequate bandwidth is available to handle all the inbound traffic generated during the test.
  • Footprint/Bandwidth Used: The bandwidth requirements are similar to many other Internet applications. The bandwidth level required would permit the test over a dial up (28.8 kbps) line for a single student. In situations where there is minimal bandwidth or wide fluctuations of bandwidth capacity for the test, administrators should prepare for the test interface to respond more slowly when moving between questions and therefore students may need more time to complete the test. For a very responsive testing environment, we recommend that you allot 56kbps of bandwidth to accommodate both upstream and downstream traffic per participant. The downstream estimate footprint for simple questions is 40-50 Kb per question. Questions with larger graphics or text passages and questions (e.g., English), can be 100 Kb or higher. Interactive items downstream estimate is between 20-70Kb. Upstream footprint for most test questions is very small, but interactive items could equal the amount of the download or greater.
  • Performance and Packet Inspection: When a web browser is used to administer a test to a student, it has to communicate with a "testing" server in a manner that is secure and in real-time. Network traffic to and from data.ncsu.edu must be allowed with NO attempt to cache its content. Firewalls, proxy servers, etc. should be configured to pass traffic on TCP ports 80 and 443 to and from data.ncsu.edu, the test server. Traffic may not appear to be web based (http and https) to some packet inspecting appliances and software.
  • Wireless Networks: There are no restrictions on the types of wired networks supported. However, wireless networks should have WEP (wired equivalent privacy) enabled on all wireless network devices in schools using wireless to access NCDPI online testing materials. Higher levels of encryption should be preferred and used over lower levels. Breaches of security of wireless networks will not affect the security of the test, but will likely reduced performance of the wireless network as a greater amount of network capacity is used for non-sanctioned activities. End-to-end security and encryption is provided through SSL (secure socket layer) encryption is deployed on the servers delivering the tests themselves.
  • Traffic Shaping Technologies: Traffic shaping is a group of techniques designed to allow prioritization policies on the transmission of data over a network link. These technologies can be used to elevate the priority of traffic between data.ncsu.edu and student workstations during test administrations. Traffic shapers would need to be set to elevate the priority of all traffic to and from data.ncsu.edu.
  • Client Workstation User Permissions: Use of NCTest should not encounter difficulty if permissions of the user account are specialized or tightly managed as long as the permissions allow normal web browsing and Flash add-on activities.
  • Viruses, Worms, Malware: The potential exists that a new virus, worm, or other malware may be released into the Internet during a period when online testing occurs. These pests exist largely to propagate themselves or take control of other computers, and they do this by infecting computers and using their processing power to transmit themselves. This process may render the computer unusable, as well as the network by clogging it with attempts to attack other computers. It is imperative that client computers and servers be maintained with up to date virus protection and operating system patches.
  • Workstation Setup Preservatives (Deep FreezeGoBackGhostClean Slate, etc.): Administrative applications that assist preservation of system state may keep software installations and/or updates from sticking. As long as these are set up not to interfere with normal web browser activities, there should be no problem with these tools and NCTest.
  • inFrame\MetaFrame\Remote Desktop: At this time we are unable to confirm if there are special considerations with regard to using remote windowing session technologies to access NCtest. Flash media components would have to be installed and proven functional on the Application/Presentation Server for students to access this test using this type of technology. The Application/Presentation Server would need to be configured to pass TCP/IP on ports 80 and 443 (encrypted) to and from data.ncsu.edu and NOT cache content across client sessions.
  • VoIP: Voice over Internet Protocol is used to replace telephone service within an organization. VoIP can create high bandwidth utilization on existing networks. Without a separate network to handle VoIP, an actively used network could reach full utilization quickly. Therefore, VoIP would have to be managed heavily or its bandwidth profile restricted during online testing events in order to archive acceptable client test session quality. VoIP typically uses 16-80 kbps per call, depending on the algorithm used to encode and compress the sound data into digital data and back.
  • Linux: At this time we are not restricting the use of Linux workstations for this project. Taking into consideration the technologies being used for development, it is possible that Linux could be used as a student environment for accessing NCTest as long as browser and Flash comparability can be locally verified as equal to what is being used in Windows and Macintosh environments.
  • Microsoft Internet Explorer version 7 Scroll Bar Stuck: Users may encounter the inability to move the bottom scroll of the left embedded flash object when zooming in but can move around with the hand tool when in a tabbed window. Cause: Taking an Online Test while in the tabbed environment will misalign the answer buttons on top the main flash object. Solution: Do not take the test in a tabbed window. (see Pop-up Policy Problem in IE7 detailed above)
Last modified: Monday, 18 April 2011, 8:58 AM