Assignment on statistics

Population is the general group of people (things, animals) where samples a are taken.

Sample is the total number of respondents that represents a population.

For example, a person wants to measure the correlation between health education and academic attainment in the state of Virginia. The population will refer to the total number of people residing within the state of Virginia that meets the requirement (e.g., elementary students, high school students, tertiary students, post-graduate students), while the sample is the small percent (e.g. 10, 20, 30)  of the the population that  meets the educational requirement of the study.

Bias is the degree of influence a researcher affects its study. This can be accessibility to a sample where it can reflect the desired outcome the researcher wants or it can also be a research apparatus that has reliability or validity problems.

4 level of measurements
Nominal is a kind of measurement where characteristics or attributes are named to classify the data.
Ordinal measurement is the level of measurement where characteristics or attributes are ranked in order as in ordinal scales.

Interval is the level of measurement where it combines both nominal and ordinal data of a sample where it computes the difference between the two forms.

Ratio is the level of measurement where there is an absolute zero. Examples of ratio scale measurements are length, temperature, speed, or weight.

ProceduresSampling typeDefinitionCharacteristicsLimitations Scenario   Convenience SamplingNon-probabilityChoosing respondents depending on the researchers accessibilityTime and cost effectiveLack of sampling accuracyPeople on street interviewsQuota SamplingNon-probabilitySampling that fills quotaTime consumingSamples cannot be isolated prior to respondents selectionResearch that needs certain numbers of respondents for various income bracketsJudgmental SamplingNon-probabilityPurpose samplingFace validity judgment on the researchers partCannot be measured or be calculated based on error of judgmentStudy on new facility of those who live on specified range within the facilitys boundarySimple Random SamplingProbabilityRandomized selection of respondents without any categoryReduces bias and estimate sampling errorsNo limitationsSurvey of county employeesStratified Random SamplingProbabilityRandom sampling that involves categorizing population members into mutually exclusive and collectively exhaustive groupsProvide more precise estimate on a heterogeneous groups, provide efficiencyDependent on variables involved in the studyDetermining average income by families in the U.S.Cluster SamplingProbabilitysampling that involves  categorizing population members into mutually exclusive and collectively exhaustive groups under defined circumstancesTime effectiveDependent on sample heterogeneityStudy attempting to measure age distribution residing in Fairfax CountySystematic SamplingProbabilityOne-stage cluster samplingSamples are chosen by intervals depending on the required variable the counting starts withEasier to draw and executeA survey of visitors a publication desk

The article  The relation between popularity and aggressive, destructive, and norm-breaking behaviors Moderating effects of athletic abilities, physical attractiveness, and prosociality  will be used to answer the succeeding questions

Sampling Procedures
The study used a portion of data from the another study, study for TRAILS (Tracking Adolescents Individual Lives Survey),  that was comprised of preadolescents. The study for TRAILS used stratified random sampling because of age exclusivity variable under 5 municipalities of northern Netherlands. A total of 122 schools participated with the total of 3,145 number of respondents where a response rate of 90.4 was reached. The baseline respondents where measured at 2,230 students and out of the baseline respondents 96.4 finished the study. Only 98 students refused to participate. Measurements on popularity and characteristics - such as norm-breaking behavior, physical aggression, destructive behavior, bullying, relational aggression, personality,physical attractiveness and athletic abilities were taken into consideration for trimming down the respondents.

Data Analysis
Statistical analyses used on the study were comprised of t-test, standard deviation (SD), analysis of variance (ANOVA), statistical significance, multi-variate correlational measurements, and the usual mean score of the population for the given characteristicsvariables. Units were measured in terms of the above stated variables (e.g., bullying, personality, etc) in reference to age bracket. The p-value differentiated from each characteristic some where compared to p .05, others to p .001, while the rest were compared to p .01. As a reference all data were plotted  clearly, as defined on the methodology, in tables and figures for easy referencing.

Findings
The study found out that variables being studied are positively correlated to how it affects them, most importantly how popularity affect the students. Though some findings when compared to the set p-values show promise some results were not that significant. Since all aspect of the variables had different comparative p-values everything was clinical and substantiated by statistical measurements. The authors were able to give discussions on the studys limitations.

0 comments:

Post a Comment