A local credit union has approached you for advice in conducting a survey among its customers. It has classified its customers, which are held alphabetically sorted in a computer file, into three categories as follows:
(i) Customers holding of active account i.e. an account that has been accessed in the past six months.
(ii) Customers holding semi active accounts i.e. an account that has been accessed in the past six to twelve months.
(iii) Customers holding dormant accounts i.e. an account that has not been accessed within the past twelve months.
Outline two random methods by which a sample might be selected. For the two methods selected, outline in detail how the sample would be selected and the merits of the method
My Answer :
Let us suppose that there are 4000 members in total.
3 Groups
GR1 – Active Accounts – 2000 members (50%)
GR2 – Semi Active Accounts – 1500 members (37.5%)
GR3 – Dormant Accounts – 500 members (12.5%)
Method 1:
Assign numbers to the names and select
GR1 – 1000 random numbers (50%)
GR2 – 562 random numbers (37%)
GR3- 63 random numbers (13%)
Total of 1625 members selected
40.6 % surveyed
Method 2:
Pseudo-random
Use a computed aided program to select 1652 names
(i) Customers holding of active account i.e. an account that has been accessed in the past six months.
(ii) Customers holding semi active accounts i.e. an account that has been accessed in the past six to twelve months.
(iii) Customers holding dormant accounts i.e. an account that has not been accessed within the past twelve months.
Outline two random methods by which a sample might be selected. For the two methods selected, outline in detail how the sample would be selected and the merits of the method
My Answer :
Let us suppose that there are 4000 members in total.
3 Groups
GR1 – Active Accounts – 2000 members (50%)
GR2 – Semi Active Accounts – 1500 members (37.5%)
GR3 – Dormant Accounts – 500 members (12.5%)
Method 1:
Assign numbers to the names and select
GR1 – 1000 random numbers (50%)
GR2 – 562 random numbers (37%)
GR3- 63 random numbers (13%)
Total of 1625 members selected
40.6 % surveyed
Method 2:
Pseudo-random
Use a computed aided program to select 1652 names