非隨機(jī)抽樣和隨機(jī)抽樣的比較ppt課件

1、Ch. 7 Selecting samplesThe need to sampleOverview of Sampling techniquesProbability samplingNon-probability samplingDefinition of termsCensus Collect and analyse data from every possible case or group memberSampling A range of methods that enable researcher to reduce the amount of data by only data

2、from a subgroup rather than all possible cases or elementsPopulation The full set of cases from which a sample is taken Figure 7.1 Population, sample and individual cases1. The need to sampleBudget constraints prevent you from surveying the entire populationTime constraints prevent you from surveyin

3、g the entire populationImpracticable to survey the entire populationYou have collected all the data but need the results quickly2. Overview of sampling techniquesProbability or representative sampling - Each case from population is known and usually is equal for all cases - (survey and experimental

4、research strategies)Non-probability or judgemental sampling - Probability of each case from the population is unknown- Impossible to answer research questions or to address objectives that require you to make statistical inferences about the characteristics of the population- (case study strategy)非隨

5、機(jī)抽樣和隨機(jī)抽樣的比較 抽樣方法作用抽樣原則誤差判斷應(yīng)用優(yōu)缺點(diǎn)非隨機(jī)抽樣研究總體的局部現(xiàn)象非隨機(jī)抽出樣本，主觀性強(qiáng)不能計(jì)算和判斷抽樣誤差可隨時(shí)隨地采用不夠科學(xué)規(guī)范，但省錢、省事、靈活方便隨機(jī)抽樣以部分推斷總體隨機(jī)抽出樣本，客觀性強(qiáng)不能計(jì)算和判斷抽樣誤差只能定期采用科學(xué)規(guī)范，但費(fèi)時(shí)、費(fèi)錢、不夠靈活方便Figure 7.2 Sampling techniques隨機(jī)抽樣非隨機(jī)抽樣簡(jiǎn)單隨機(jī)抽樣系統(tǒng)抽樣分層抽樣分群抽樣多步抽樣配額抽樣雪球抽樣便利抽樣自選抽樣目的抽樣極端抽樣同質(zhì)抽樣不均勻抽樣典型抽樣關(guān)鍵抽樣3. Probability samplingProcess of probability sa

6、mpling(2) decide on a suitable sample size(3) select the most appropriate sampling technique and select the sample(4) check the sample is representative of the population(1) identify a suitable sampling frame based on your research questions or objectivesSample frame A complete list of all the cases

7、 in the population from which your sample will be drawnSample size the number of cases used for the research analysisStatistical inference a probable conclusion about a population on the basis of data of sample Law of large number Larger sample size can better represent the population than Smaller s

8、ample sizeHow to choose the sample size?The confidence you need to have in your data: the level of certainty that the sample can represent the total population(confidencesample size)The margin of error that you can tolerate: the accuracy you require for any estimates made from your sample (accuracy

9、sample size)The types of analysis you will undertake: (Categoriessample size; minimum threshold of each technique)The size of total population from which your sample is being drawnResponse rateReasons of non-response: Unreachable; ineligible; inability；refusal; total number of responsesTotal Respons

10、e rate = -total number in sample - ineligible total number of responsesActive Response rate = - total number in sample (ineligible + unreachable)Population, sampling frame, samplesSelect appropriate sampling techniqueFive main sampling techniques(1) simple random (2) systematic(3) stratified random(

11、4) cluster(5) multi-stagesampling technique(1)Simple random sampling(a) Number each case in your sampling frame with a unique number(b) Select cases using random numbers until your actual sample size is reached (pp218;587 for “Random number tables).sampling technique(2)Systematic sampling(a) Number

12、each case in your sampling frame at regular intervals(b) Select the first case using a random number(c) calculate the sampling fraction (抽樣比) (d) select subsequent cases systematically using the sampling fraction to determine the frequency of selection actual sample sizeSampling fraction = - total p

13、opulation Sampling fraction: The proportion of the total population that you need to select.1. Decide on sample size: n2. Divide frame of N individuals into n groups of k individuals: sampling fraction k=n/N3. Randomly select one individual from the 1st group 4. Select every 1/k-th individual therea

14、fterSystematic SampleN = 64n = 81/k = 8First Groupsampling technique(3) Stratified random sampling strtifaid (a) choose the stratification variable(s)(b) divide the sampling frame into the discrete strata(c) number each of the cases within each stratum with a unique number(d) select your sample usin

15、g either simple random or systematic samplingStratified Sample1. Divide Population into SubgroupsMutually ExclusiveExhaustiveAt Least 1 Common Characteristic of Interest2. Select Simple Random Samples from SubgroupsAll StudentsPart-timeFull-timeSamplesampling technique(4)Cluster sampling(a) Choose t

16、he cluster grouping for your sampling frame(b) number each of the clusters with a unique number. The first cluster is numbered 0, the second 1, and so on(c) select your sample using some form of random sampling Cluster Sample1. Divide Population into ClustersIf Managers are Elements then Companies a

17、re Clusters2. Randomly Select Clusters3. Survey All or a Random Sample of Elements in ClusterCompanies (Clusters)Samplesampling technique(5) Multi-stage samplingOverview of probability sampleQuota sampling Purposive samplingSnowball samplingSelf-selection samplingConvenience sampling4. Non-probabili

18、ty sampling4. Non-probability sampling(1) Quota sampling(a) divide the population into specific groups(b) calculate a quota for each group based on relevant and available data(c) give each interviewer an assignment, which states the number of cases in each quota from which they must collect data(d)

19、combine the data collected by interviewer to provide the full sampleSamples4. Non-probability sampling2 Purposive sampling (judgemental sampling) Use researchers judgement on sampling(a) extreme case sampling extreme case will be relevant in understanding and explain more typical cases. E.g. study o

20、n excellent students (b) heterogeneous sampling complete different cases, maximum variation will be particular interest and value and will represent the key themes. E.g. study all special students (c) homogeneous sampling enable you to study the group in great depth. E.g. study on all stu. with IELT

21、S 6.0. (d) critical case sampling if it happens in one critical case, can it happens to everyone. E.g. study on a successful stu. with low entrance grade. (e) typical case sampling illustrate a profile with a representative case. E.g. study a normal stu. with average study performance 4. Non-probability sampling3 Snowball sampling (a) make