System Life Cycle
System Life Cycle
● A system analysis team is often brought in to review an existing system and suggest several
improvements
● This existing method used may be either a manual paper based system or a computer based operation
that is no longer regarded as adequate for the task
Interviews
● Involves a one-to-one question and answer session between the analyst and the employee
● Advantages
○ It gives the opportunity to motivate the interviewee into giving open and honest answers to the
analyst’s questions
○ It allows the analyst to probe for more feedback from the interviewee, as it is easier to extend a
question
○ It is possible to modify questions as the interview proceeds and ask questions specific to the
interview
○ It is a good method if the analyst wants to probe deeply into one specific aspect of the existing
system
● Disadvantages
○ It can be time consuming, relatively expensive and interviewee cannot remain anonymous
Questionnaires
● Involves sending out questionnaires to the work force and to the customers to find out views of existing
system and find out some key tasks that are carried out
● Advantages
○ The questions can be answered quite quickly
○ It is a relatively inexpensive method
○ Individuals can remain anonymous if they want
○ It allows quick analysis of the data
● Disadvantages
○ Often the number of returned questionnaires is low
○ The question are rather inflexible since they have to be generic
○ There is no immediate way to clarify vague or incomplete answer to a question
Verification
● A way of preventing errors when data is copied from one medium to another
● Double entry
○ Data is entered twice using 2 different people
○ The computer compares the 2 entries, during or after data entry and identifies any differences
● Visual check
○ Checking for errors by comparing entered data on the screen with the data in the original
document
Validation
● Process where data is checked to see if it satisfies certain criteria when input into a computer for
example to see if the data falls within accepted boundaries
● Range check
○ Checks whether data is within given/acceptable values
○ E.g. a person's age should be in range more than 0 and less than 150
● Length check
○ Checks if the input data contains the required number of characters
○ E.g. if a field needs six digits then inputting a five or seven digit number causes an error
● Character/type check
○ Checks that the input data does not contain invalid characters
○ E.g. a person’s name should not contain any numbers but a person’s height should only contain
digits
● Format/picture check
○ Check that data is in a specific format
○ E.g. date should be in form dd/mm/yy
● Limit check
○ Similar to range check except that only one of the limits is checked
○ E.g. input data must be less than 10
● Presence check
○ Check if data is actually present and has not been missed out
○ E.g. in an electronic form, a person's telephone number may be required field and if no data is
present this should give rise to an error message
● Consistency check
○ Checks if fields correspond with each other
○ E.g. if ‘Mr’ typed into a field called title then the gender field must contain either ‘M’ or ‘Male’
● Check digit
○ Looks at an extra digit which is calculated from the digits of a number and then put on the end of
the number
○ E.g. check digits can identify 3 types of errors
■ If 2 digits have been inverted during input
■ An incorrect digit entered twice
■ A digit missed out altogether
Testing strategies
● Testing of each module needs to be done to ensure each one functions correctly on its own
● Once the development of each module is completed, the whole system needs to be tested (i.e. all
modules functioning together)
● Even though each individual module may work satisfactorily, when they are all put together there may
be data clashes, incompatibility and memory issues
● All of this may lead to a need to improve the input and output methods, file and database structures,
validation and verification methods
● It is a very time-consuming process but the system must be as perfect as possible before it goes live
● Testing will use many different types of data, which will fall into one of three categories: normal, extreme
or abnormal
○ Normal
■ This data which is acceptable/valid and has an expected outcome
■ E.g. the month can be any whole number in range of 1-12
○ Extreme
■ This is data at the limits of acceptability/validity
○ Abnormal
■ This is data outside the limits of acceptability/validity and should be rejected or cause an error
message