165
Psychology (Subject) / Research Methods (Lesson)
Advertisement
- Sampling plan > specifies how the respondent will be selected > this is one way to overcome the “paradox of sampling” by consideration or the procedure by which the sample is obtained > consideration that the representativeness is not a property of the sample but rather of the procedure by which the sample is obtained
- Point Estimates - central values in frequency distributions - mean, median, mode -
- Interval Estimates - Variability in frequency distributions - frequently expressed in terms of the margin of error - margin of error is conceptually related to the standard error: > indictated the degress of imprecision with which we have estimated
- Bias / Stability- precision - Two forms of deviation from the true population value: > Bias: the difference between the true population value and our estimate of it from a sampling distribution > Stability/ precision: the variability of observations or samples drawn on the basis of probability sampling > Instability results when the observations within a sample are highly variable and the number of observations is small. The more homogenous the members of the population, the fewer of them need to be sampled
- Simple random sampling > the sample is selected from an unidivided populations (simple) and chosen by a process that gives every unit in the population the same chance of being selectes (random).
- Improving Accuracy in Random Sampling - stratified random sampling - stratified random sampling: 1) we divide the population into a number of parts 2) and randomly sample in each part independently (unbiased procedure) - It can pay off in an improved accuracy by enabling researchers to randomly sample strata that are less variable than the original population
- Confidence Intervals - increasing the confidence level from 90,95,99% will also increase the interval between the upper and lower limits - if we increase the sample size, the confidence interval gets smaller > Bayesian approach: we can be 95% confident, that the population value we are trying to estimate falls between the upper and lower limits > Frequentist approach : “With repeated samplings for each sample of % confidence intervals, we will be correct in % of our samples when we say that the quantity estimated will fall within the % confidence interval''
- Area probability sampling (other selection procedure) - population is divided into selected units that have the same probability of being chosen as the unselected units in the population cluster > multistage cluster sampling
- Systematic sampling (other selection procedure) - involves the methodological selection in sequences seperated on lists by the interval of selection 1) a random start 2) the sampling interval > sample may not be exactly random
- Haphazard or fortuitious samples (other selection procedure) - nonrandom strategy of selection - a poll by the media where people are urged to call and vote
- Advertisement
- Quota sampling (other selection procedure) - involves obtaining specified numbers of respondents to create a sample that is roughly proportional to the population - characteristics of the desired respondents are spelled out in advance; asily biased
- Nonresponse Bias > increasing the effort to recruit the nonrespondents decreases the bias of point estimates in the sample, independent on what design you are using > prepaid monetary incentives produce higher response rates than promised incentives or gifts offered with the initial encounter > a proposed theoretical model was based on Wald’s “model of nonresponse''
- Volunteer subjects - might differ from non volunteers - comparison strategies (volunteers vs nonvolunteers) 1) archive containing information about all subjects, volunteering and non volunteering subjects > compare both groups 2) college class ex.: volunteers are recruited, nonvolunteers are identified from the class rooster > test is given to all later (by different person) > differences are compared - Volunteering might have both general and specific predictors - characteristics: (in descending order) 1) educated 2) higher social class 3) intelligent 4) approval-motivated 5) sociable - correlated of volunteering (in descending order) 1) subject interest 2) expectations of favourable evaluations 3) task importance 4) guilt, happiness, and confidence 5) material incentives
- Missing Data - the primary problem of missing data is the introduction of bias into our estimates - decreased statistical power MCAR- Missing completely at random: said to occur wgen the missingness is unrelated to any variable of substantive interest (unbiased) MAR - Missing at random: said to occurs when the missingness is related to the variables of interest, but when these relationships can be accounted for by other variables that have been observed (bias can be reduced) MNAR - Missing not at random: said to occur when missingness is related to variables of interest and cant be accounted for by others (biased)
- Procedures for dealing with missing data - Nonimputational procedures: those that yield estimates of parameters without fillinf in the missing data Listwise deletion: drop all subjects that have missing data Pairwise deletion: does not drop any subjects but simply computes the parameters on all data that are available Maximum likelhood estimation: complex mathematical computations Bayesian estimation: '' - Putational procedures: those that yield their estimates after filling in the missing data Single imputation: each missing value is replaced by a reasonable estimate of what the value might habe been had it been observed Multiple imputation: each missing value is not replaced by a single estimate but by a set of m reasonable estimates that will yield m pseudocomplete datasets