Contents

General Information

   Features

   About/References

   Situational recommendations

Main interface

   Overview

   Demographic Info

   Data Recording

   Statistics

   Graphs

   Cursor Event Playback

 

What does iCara do?
iCara is a basic usability testing tool designed for covert data recording of user interface activity within controlled laboratory-based usability research. As a local-PC installed program, iCara does not require remote server internet-based services to record user activity. In addition, iCara provides a means by which user activity data can be exported to Microsoft Excel or some other Excel compatible spreadsheet program (such as OpenOffice Calc) for the purpose of analyzing and graphing data at more advanced levels.

SYSTEM REQUIREMENTS
iCara has been tested on Windows XP, and Windows 7. Although it has not been tested on Windows Vista or 98, it should be wholly compatible. 256MB RAM; 400MHZ CPU minimum.

*ATTENTION Windows 7 users:
You will need to run iCara in 'Administrator' mode. Right-click the iCara icon and choose 'Run as Administrator' from the right-click menu if it does not launch when clicked normally.
 

Features
*Local PC install; no remote server needed
*Record total mouse distance traveled in pixels and meters for entire session
*Calculate cursor rate in pixels/second and meters/second
*Calculate time-on-task *Record Left/Right mouse click count
*Record XY on-screen position of all clicks
*Record XY position of cursor coordinates (~10 samples per second)
*Calculate mean of X values *Calculate variance of X values *Calculate standard deviation of X values
*Calculate mean of Y values
*Calculate variance of Y values
*Calculate standard deviation of Y value
*4 user-defined fields for demographic data entry
*Real-time playback of cursor movement and click events on actual interface
*Real-time playback of cursor movement and click event scatter-plotting on custom crosshair click map interface
*Separate data files for click event X and Y data, X and Y statistics, X and Y cursor position data
*Combined XY cursor-position-over-time data graphing *Unique X value cursor-position-over-time data graphing
*Unique Y value cursor-position-over-time data graphing
*Automatically writes all mouse data and statistics to text-file comma-delimit format for import into any data analyses program (e.g. SPSS, SAS)
*Option to export all mouse data to Excel.xls spreadsheet format (requires Microsoft Excel)
Top


About
iCara is an interface design assessment tool for use in controlled laboratory testing environments. The iCara program is not an
internet-based program and therefore requires no subscription. iCara is a locally installed program that can provide testing data for
WinForms software programs that are not web-based. However, iCara can record user any type of activity on a PC interface; including
internet activity/surfing and/or proprietary web-based program evaluation.


References
Nakamichi, N., Shima, K., Sakai, M., and Kenichi, M., (2006). Detecting low usability web pages using quantitative data of users' behavior.
Proceedings of 28th International Conference on Software Engineering (ICSE'06), 569-576.
Top


Situational recommendations
It is recommended that care be taken when using iCara with participants so as not to divulge the covert control of the program.
The following recommendations can be helpful while using iCara within testing sessions:
1) Consider using an external monitor placed back-to-back with the participant monitor so as to view real-time activity during the
measured session; this will allow any unusual activity to be viewed and demarcated as the session progresses
2) Consider using a separate keyboard along with the external 'experimenter' monitor so as to control the iCara from the secondary location
3) Consider turning the participant monitor ON and OFF as needed. This will assure that any on-screen activity or interaction with the iCara
program by the experimenter remains private. This can be done via the power button on an external monitor; on laptop PC's this can be accomplished via PC settings software.
Top


Demographics Info
Allows identifying information of session participant to be entered in the UDF 'user-defined-fields'. All UDF items must contain data in order for recording to be activated.

Data Recording
Records cursor activity on the user interface at 10 samples per second. All demographic information on the Demographic page must be completed for record to commence. The recording process generates the data by which all statistics for X and Y cursor values are calculated. In addition, there is a feature called 'turn OFF statistics calculation' can that can be ticked so as to prevent the calculation of statistics upon stopping of the recording process. This feature is handy for long recording sessions being that the calculation of statistics can last 30 to 60 seconds for long sessions. The Graphs page provides the means to download all cursor data to an .xls file extension; statistics can then be easily calculated via Excel on the raw data.


Statistics
Descriptive statistics for cursor X and Y data are calculated upon termination of recording. This feature may be turned off by ticking the 'turn OFF statistics calculation' box. As mentioned above, this feature is handy for long recording sessions being that the calculation of statistics can last 30 to 60 seconds for long sessions. Again, the Graphs page provides the means to download all cursor data to an .xls file extension; statistics can then be easily calculated via Excel on the raw data.


Graphs
The graphing of the cursor data is as follows: 1) 'All X, Y Values' graph. Graphs the combination of both X and Y cursor coordinate positioning at 10 samples per second This graph includes redundant X and Y coordinate position values as well as all unique X and Y coordinate position values. 2) 'Unique X Values' graph. Graphs only the unique X axis (horizontal) cursor values. For example, if the cursor records an X axis position value of 935 eighty-nine times within the data set, this X = 935 data point would only be plotted once on this graph. 3) 'Unique Y Values' graph. Similar to the 'Unique X Values' graph, this graph contains only the unique Y axis (vertical) cursor values. For example, if the cursor records a Y axis position value of 514 twenty-seven times within the data set, this Y = 514 data point would only be plotted once on this graph.


Playback
Allows for the playback of recorded cursor movement. These cursor movement files can be played back on the same interface where recorded, or the option to launch the 'Click Plot' for playback may be chosen. The 'click plot' option will launch a blank white screen that is equipped with a crosshair with red and blue lines separating the screen into 4 quadrants; the click plot screen is also sensitive to left and right mouse clicks and will record a black dot where these clicks occur on the click plot screen. The end result is a scatterplot of click activity that allows the analysis of click clustering to be made visually.

Top

Demographics
Allows user-defined case identifying data to be entered for each recording session.
 


Top


Data Recording
Recording cursor activity on the user interface:
1) All demographic information on the Demographic Information page must be completed first
2) Decide if you wish to have iCara automatically calculate statistics to be displayed on the 'Statistics' interface after recording is stopped*
3) Click the green 'Start' button
4) Begin moving the cursor around and/or generate left-clicks and right-clicks on the screen
5) Click the red 'Stop' button to stop recording
*large datasets may require 30-45 seconds to calculate all statistics
Note: upon clicking the 'Start' button the mouse cursor on-screen will suddenly 'jump' or move to the top left corner of the screen; this assures that each recording of cursor activity has an identical starting point


Hidden Mode Keyboard Shortcuts
The following keyboard shortcuts allow you to hide/unhide iCara, start/stop recording while hidden, and exit (kill program) during hidden mode:
       Hide = H + F12
       Unhide = U + F12
       START recording while hidden = Alt + F12
       Stop recording while hidden = Ctrl + F12
       Exit (kill program) = X + F12
 


Top


Statistics
Statistics are calculated automatically unless the override is checked on the statistics page. In addition, all statistics are sent to a file named
'StatsData.txt' located within the 'data' folder. The statistics within this txt file are also preceded by the demographic information associated
with the file, datetime stamp, and the screen resolution at the time the statistics were generated.
 


The design intent of iCara was to fill a void that occurs within a number of other interface recording programs. This void primarily consists of the calculation of the statistical parameters surrounding cursor distance traveled and the rate of cursor movement. Poorly designed interfaces can be associated with interfaces greater user interface activity with respect to cursor rate and cursor distance traveled (see Nakamichi, N., Shima, K., Sakai, M., and Kenichi, M., 2006). An analogy of the importance of such an association is as follows:
 

Imagine if a large retail store wanted to determine how easily a customer could find a specific item in their store...
1) Imagine that the floor of the store is the 'interface', and the human participant is the mouse cursor
2) The experimental design is a paradigm where participants start at the front of the store, and each participant is required to walk the store and locate a specific item
3) Imagine that the experiment does not record how far each participant has walked, nor how fast each participant walked during the task; they only record the amount of 'time' it took the participant to find the target item.
 

What if the participant stood still for 30-40 seconds on two or three occasions looking around or reading the aisle content signs? Such activity would not be accounted for in the data set; few would argue it fails to contribute to insight regarding the research objective. In such an experimental paradigm there would be an optimal number of 'turns' that each participant could make to get to the correct location in the most efficient manner possible; each turn down a wrong aisle a participant made is not recorded under this paradigm---and each wrong turn down an aisle adds both 'distance' and time to the overall participant task--both critical factors in recording 'rate'. Even if wrong turns are recorded, the overall distance traveled is still missing as a critical component of the task. Lacking a means by which to record distance traveled and participant a subsequent rate, leaves time-on-task as the lone data point in determining how long the task of locating the item took. Two participants could have the exact same time-on-task to locate the same item, but one participant could have taken a greater number of wrong turns thus requiring them to increase their speed knowing that they had covered an unnecessary amount of distance via their wrong turns.
Top


Graphs
The iCara program creates 3 graphs of cursor data inclusive of the following:
*Combined XY cursor coordinates over time of task (10/sec. sample rate)
*Unique X and unique Y cursor position values
       -all redundant X and Y values removed
 

The following click-associated data is also recorded on the Graphs interface:
*Click ID: 1 = Left Click; 2 = Right Click
*X and Y position of click when clicks are rendered on interface

The iCara graphed data merely presents the raw data which can be exported into Microsoft Excel. The exporting process
to Excel can allow for far greater customization of the graphs; graphed data within iCara merely provides a basic visual
representation of the data generated from the session.
 


Top


Cursor Event Playback
iCara allows for the playback of all cursor movement generated during a recording session. This cursor playback can
be accomplished on an actual interface, or it can be accomplished by playing back the cursor movement on the
click-event-mapping interface. The click event mapping interface (see below) is sensitive to mouse clicks and will
place a black dot at the position of all clicks. The resulting 'scatterplot' can provide quick insight into clustered click
activity on an interface. All cursor data files are saved as .dat file extension files within the 'logs' file folder.

 



 

Top