Output Design
Output is the primary purpose of any system.
These guidelines apply for the most part to both paper
and screen outputs. Output design is often discussed before
other aspects of design because, from the
client's point of view, the output is the system. Output is what the
client is buying when he or she pays for a development project.
Inputs, databases,
and processes exist to provide output.
-
Problems often associated with business information output are information delay,
information (data) overload, paper domination, excessive distribution, and nontailoring.
-
Mainframe printers: high volume, high speed, located in the data center
Remote site printers: medium speed, close to end user.
-
COM is Computer Output Microfilm. It is more compact than traditional output and
may be produced as fast as non-impact printer output.
-
Turnaround documents reduce the cost of internal information processing by reducing
both data entry and associated errors.
-
Periodic reports have set frequencies such as daily or weekly; ad hoc reports are
produced at irregular intervals.
-
Detail and summary reports differ in the the former support day-to-day operation
of the business while the latter include statistics and ratios used by managers to
assess the health of operations.
-
Page breaks and control breaks allow for summary totals on key fields.
-
Report requirements documents contain general report information and field specifications;
print layout sheets present a picture of what the report will actually look like.
-
Page decoupling is the separation of pages into cohesive groups.
-
Two ways to design output for strategic purposes are (1) make it compatible with
processes outside the immediate scope of the system, and (2) turn action documents
into turnaround documents.
-
People often receive reports they do not need because the number of reports received
is perceived as a measure of power.
-
Fields on a report should be selected carefully to provide uncluttered reports,
facilitate 80-column remote printing, and reduce information (data) overload.
-
The types of fields which should be considered for business output are: key fields
for access to information, fields for control breaks, fields that change, and exception
fields.
-
Output may be designed to aid future change by stressing unstructured reports,
defining field size for future growth, making field constants into variables, and
leaving room on summary reports for added ratios and statistics.
-
Output can now be more easily tailored to the needs of individual users because
inquiry-based systems allow users themselves to create ad hoc reports.
-
An output intermediary can restrict access to key information and prevent unauthorized
access.
-
An information clearinghouse (or information center) is a service center that
provides consultation, assistance, and documentation to encourage end-user development
and use of applications.
-
The specifications needed to describe the output of a system are: data flow diagrams,
data flow specifications, data structure specifications, and data element specifications.
Output Documents
Printed Reports
- External Reports: for use or distribution outside the organization; often on preprinted forms.
- Internal Reports: for use within the organization; not as "pretty", stock paper, greenbar, etc.
- Periodic Reports: produced with a set frequency (daily, weekly, monthly, every fifth Tuesday, etc.)
- Ad-Hoc (On Demand) Reports: irregular interval; produced upon user demand.
- Detail Reports: one line per transaction.
- Summary Reports: an overview.
- Exception Reports: only shows errors, problems, out-of-range values, or
unexpected conditions or events.
Input Design
-
A source document differs from a turnaround document in that the former contains
data that change the status of a resource while the latter is a machine readable
document.
-
Transaction throughput
is the number of error-free transactions entered during a specified time period.
-
A document should be concise because longer documents contain more data and so
take longer to enter and have a greater chance of data entry errors.
-
Numeric coding substitutes numbers for character data (e.g., 1=male, 2=female);
mnemonic coding represents data in a form that is
easier for the user to understand and remember.
(e.g., M=male, F=female).
-
The more quickly an error is detected, the closer the error is to the person who
generated it and so the error is more easily corrected.
-
An example of an illogical combination in a payroll system would be an option to
eliminate federal tax withholding.
-
By "multiple levels" of messages, Shneiderman means
allowing the user to obtain more
detailed explanations of an error by using a help option, but not forcing a lengthy
message on a user who does not want it.
-
An error suspense record would include the following fields: data entry operator
identification, transaction entry date, transaction entry time, transaction type,
transaction image, fields in error, error codes, date transaction reentered successfully.
-
A data input specification is a detailed description of the individual fields
(data elements) on an input document together with their characteristics (i.e., type
and length).
Error Messages to be displayed
for the end user
- Be specific and precise, not general, ambiguous, or vague.
(BAD: Syntax error, Invalid entry, General Failure)
- Don't JUST say what's wrong---- Be constructive; suggest what needs to be done to correct
the error condition.
- Be positive; Avoid condemnation. Possibly even to the point of avoiding pejorative
terms such as "invalid" "illegal" or "bad."
- Be user-centric and attempt to convey to the user that he or she is in control by
replacing imperatives such as "Enter date" with wording such as "Ready for date."
- Consider multiple message levels: the initial or default error message can be brief
but allow the user some mechanism to request additional information.
- Consistency in terminology and wording.
- place error messages in the same place on the screen
- use consistent display characteristics (blinking, color, beeping, etc.)
Interactive Screen Design
-
The primary differences between an interactive and batch environment are:
- interactive processing is done during the organization's prime work hours
- interactive systems usually have multiple, simultaneous users
- the experience level of users runs from novice to highly experienced
- developers must be good communicators because of the need to design systems with
error messages, help text, and requests for user responses.
-
The seven step path that marks the structure of an interactive system is
- Greeting screen (e.g., company logo)
- Password screen -- to prevent unauthorized use
- Main menu -- allow choice of several available applications
- Intermediate menus -- further delineate choice of functions
- Function screens -- updating or deleting records
- Help screens -- how to perform a task
- Escape options -- from a particular screen or the application
-
An intermediate menu and a function screen differ in that the former provides choices
from a set of related operations while the latter provides the ability to perform
tasks such as updates or deletes.
-
The difference between inquiry and command language dialogue
modes is that the
former asks the user to provide a response to a simple question (e.g., "Do you really
want to delete this file?") where the latter requires that the user know what he
or she wants to do next (e.g., MS-DOS C:> prompt; VAX/VMS $ prompt; Unix shell prompt).
GUI Interface (Windows, Macintosh) provide Dialog Boxes to prompt user to input required
information/parameters.
-
Directions for designing form-filling screens:
- Fields on the screen should be in the same sequence as on the source document.
- Use cuing to provide the user with information such as field formats (e.g., dates)
- Provide default values.
- Edit all entered fields for transaction errors.
- Move the cursor automatically to the next entry field
- Allow entry to be free-form (e.g., do not make the user enter leading zeroes)
- Consider having all entries made at the same position on the screen.
-
A default value is a value automatically supplied by the application when the user
leaves a field blank. For example, at SXU the screen on which student names and
addresses are entered has a default value of "IL" for State since the majority of
students have addresses in Illinois. At one time "312" was a default value for Area Code,
but with the additional Area Codes now in use (312, 773, 708, 630, 847) providing a default
value for this field is no longer as useful.
-
The eight parts of an interactive screen menu are:
- Locator -- what application the user is currently in
- Menu ID -- allows the more experienced user access without going through the entire menu tree.
- Title
- User instructions
- Menu list
- Escape option
- User response area
- System messages (e.g., error messages)
-
Highlighting should be used for gaining attention and so should be limited to critical
information, unusual values, high priority messages, or items that must be changed.
-
Potential problems associated with the overuse of color are:
- Colors have different meanings to different people and in different cultures.
- A certain percentage of the population is known to have color vision deficiency.
- Some color combinations may be disruptive.
-
Information density is important because density that is too high makes it more
difficult to discern the information presented on a screen, especially for novice
users.
-
Rules for defining message content include:
- Use active voice.
- Use short, simple sentences.
- Use affirmative statements.
- Avoid hyphenation and unnecessary punctuation.
- Separate text paragraphs with at least one blank line.
- Keep field width within 40 characters for easy reading.
- Avoid word contractions and abbreviations.
- Use nonthreatening language.
- Avoid godlike language.
- Do not patronize.
- Use mixed case (upper and lower case) letters.
- Use humor carefully.
-
Symmetry is important to screen design because it is aesthetically pleasing and
thus more comforting.
-
Input verification is asking the user to confirm his or her most recent input
(e.g., "Are you sure you want to delete this file?")
-
Adaptive models are useful because they adapt to the user's experience level as
he or she moves from novice to experienced over time as experience with the system
grows.
-
"Within User" sources of variation include: warm up, fatigue, boredom, environmental
conditions, and extraneous events.
-
The elements of the adaptive model are:
- Triggering question to determine user experience level
- Differentiation among user experience
- Alternative processing paths based on user level
- Transition of casual user to experienced processing path
- Transition of novice user to experienced processing path
- Allowing the user to move to an easier processing path
-
Interactive tasks can be designed for closure by providing the user with feedback
indicating that a task has been completed.
-
Internal locus of control is making users feel that they are in control of the
system, rather than that the system is in control of them.
-
Examples of distracting use of surprise are:
- Highlighting
- Input verification
- Flashing messages
- Auditory messages
-
Losing the interactive user can be avoided by using short menu paths and "You
are here" prompts.
-
Some common user shortcuts are: direct menu access, function keys, and shortened
response time.
|