Chapter 7 - methedology

4 levels of nonreponse

1) Complete refusal: none of the questions were answered. 2) Break off: less than 50% of the questions were answered. 3) Partial response: 50-80% of all questions were answered. 4) Complete response: over 80% of all questions were answered.

Need to sample

1) Higher overall accuracy. 2) More time for design and piloting 3) Collecting more detailed information.

Non-probability sampling

1) Quota Sampling: used for interviews. - Divide the population into specific groups. - Calculate a quota for each group based on relevant data. - Give each interviewer an assignment. - Combine the data - full sample. 2) Purposive Sampling: use your judgement to select cases that will answer your questions. You include and exclude cases and therefore it does not represent the target population. - Extreme case: focuses on unusual cases. The outcome will answer the question. You can use that to understand and explain other cases. - Heterogeneous: uses your judgement to choose diverse participants to have maximum variation. - Homogenous: focuses on subgroups which all members are similar. - Critical Case: select important cases so you understand. - Typical Case: used to provide a profile. - Theoretical Case: special case to build theory-grounded theory. 3) Volunteer sampling: participants are voluntarily part rather than chosen. - Snowball sampling: used when it is difficult to identify members. You make contact with one or two cases of the population. You ask these cases to identify other cases and you stop when no new cases are given or available. - Self-selection: allow each case to identify their desire to take part in the research. You publicise your need for cases. 4) Haphazard Sampling: occurs when case are selected because they are available. - Convience: selecting cases haphazardly because they are available.

Probability Sampling

1) Simple random sampling: select the sample at random from the sampling frame using a computer or random number tables. Number each of the cases with a unique number. First=0, Second=1. Select cases using random numbers until size is reached. 2) Systematic Random Sampling: select the sample at regular intervals. Number each of the cases with a unique number. First=0, Second=1. Select first case using a random number. Calculate the sampling frame. Select cases systematically using the sampling fraction to determine the frequence of selection. 3) Stratified random sampling: divide target population into two or more relevant strata based on one or more attributes. Choose stratified variable. Divide sampling frame into different strata. Number each strata with a unique number. First=0, Second=1. Select your sample using either simple or systematic random sampling. 4) Cluster sampling: divide target population into descrete groups. Groups=clusters. Choose the cluster grouping for your sampling frame. Number each cluster with a unique number. First=0, Second=1. Select your sample cluster, using it randomly. 5) Multi-method Sampling: development of cluster sampling. Used to overcome problems associated with a geogrpahically population when face-to-face contact is needed. Or when it is expensive or time consuming.

Sampling Frame

A complete list of all the cases in the target population from which your sample will be drawn.

A sample size of 30 is a good rule for enabling statistical analysis to be undertaken.

A sample size of 30 is a good rule for enabling statistical analysis to be undertaken.

Snowball Sampling

Access difficult or hidden populations. Overcoming the problem of not having a sampling frame. Theorise inductively in a qualitative study.

Representativeness

Comparing data generated from your sample with know characteristics of your population.

Probability Sampling

Every case has a known and equal chance of selection.

Multi-method Sampling

development of cluster sampling. Used to overcome problems associated with a geogrpahically population when face-to-face contact is needed. Or when it is expensive or time consuming.

Cluster Sampling

divide target population into descrete groups. Groups=clusters. Choose the cluster grouping for your sampling frame. Number each cluster with a unique number. First=0, Second=1. Select your sample cluster, using it randomly.

Stratified Random Sampling

divide target population into two or more relevant strata based on one or more attributes. Choose stratified variable. Divide sampling frame into different strata. Number each strata with a unique number. First=0, Second=1. Select your sample using either simple or systematic random sampling.

Haphazard Sampling

occurs when case are selected because they are available. - Convience: selecting cases haphazardly because they are available.

Volunteer Sampling

participants are voluntarily part rather than chosen. - Snowball sampling: used when it is difficult to identify members. You make contact with one or two cases of the population. You ask these cases to identify other cases and you stop when no new cases are given or available. - Self-selection: allow each case to identify their desire to take part in the research. You publicise your need for cases.

Simple Random Sampling

select the sample at random from the sampling frame using a computer or random number tables. Number each of the cases with a unique number. First=0, Second=1. Select cases using random numbers until size is reached.

Systematic Random Sampling

select the sample at regular intervals. Number each of the cases with a unique number. First=0, Second=1. Select first case using a random number. Calculate the sampling frame. Select cases systematically using the sampling fraction to determine the frequence of selection.

Purposive sampling

use your judgement to select cases that will answer your questions. You include and exclude cases and therefore it does not represent the target population. - Extreme case: focuses on unusual cases. The outcome will answer the question. You can use that to understand and explain other cases. - Heterogeneous: uses your judgement to choose diverse participants to have maximum variation. - Homogenous: focuses on subgroups which all members are similar. - Critical Case: select important cases so you understand. - Typical Case: used to provide a profile. - Theoretical Case: special case to build theory-grounded theory.

Quota Sampling

used for interviews. - Divide the population into specific groups. - Calculate a quota for each group based on relevant data. - Give each interviewer an assignment. - Combine the data - full sample.