Information Handling - Essay Example

There are key terms used in sampling.  They include: -(i) Population – This refers to the set of units under study.  The study should not go beyond the population i.e. outside the population. (Abell, & Oxbrow, 2001.p. 200).(ii) Sample - This refers to a part of the population.  It is usually hard to study the population as a whole hence the use of the sample (Grimshaw, 2003p.86).Basically, a sample should reflect the population (Taylor, Farrell, 1994p. 118).  There are ways in which a stratified sample can reflect a population.

  For example in our case of Tundra.com, the stratified sample of size 300 can reflect the population (David P. Best (1996.p166p).  This is possible in ways such as: -(i) Using the proportion of customers given for each group in relation to the total proportion.  Multiplying the total sample size with the proportion for each group will give the sample size.(ii) Using the variability of expenditure for each group we can compute sample proportion.  This is through dividing the variability of expenditure for each group by the total variability and multiplying by the sample size of 300.(iii) Using the cost per respondent for each group with regards to gathering data can help in giving sample size.

Due to the need of Tundra.com to break into the higher education market, an interview has been carried out.  This was aimed at gauging the viability of the market.  It was carried out in London.. It was carried out in London.SummaryDifferent ways were used in finding how a 300 size sample could reflect the population.Other types of stratified random sampling include: -(i) Proportional - This is where samples are taken in proportion to the population.Advantages It is precise.It is clearly representative.

DisadvantagesAssumes uniformity which is sometimes unrealistic.It requires knowledge which might lack. (ii) Disproportionate - This is where sample is taken without consideration of the population size.AdvantagesUseful where costs of collecting data differ among subgroups.Helps where different responses from different strata of people are expected.Disadvantages It's abstract in terms sample size determination.Has no uniformity.(iii) Optimal - This is where sampling is made to yield the least attainable variability.

AdvantagesEmphasises optimum allocation of units.It convenient and time saving due to picking of sample units.DisadvantagesIt's ideal because the specific units may not be readily available.Representativeness of Samples(i) Sample as per proportion of customersThis sample seems a bit uniformly spread since it has a smaller range compared to the others. The range is (90 - 18) = 72(ii) Sample as per variability of expenditureThis sample seems to be a poor representative of the population. It is affected by extreme values.

It has also a large range of (104 - 13) = 91(iii) Sample as per data gathering cost per respondentThis sample seems to give a dismal result to reflect the population. It has the largest range comparatively i.e. (135 - 13) = 120Recommendation on Stratified Sampling MethodsThe stratified random sampling methods which include proportion, optimal and disproportiate methods