HMPYC80 ASSIGNMENT 3 2022

HMPYC80 ReseaRCH MetHodologY (2022 - seMesteR 1 - assignMent 3) HMPYC80 ASSIGNMENT 3 SEMESTER 1 YEAR 2022 Page 1 of 14 HMPYC80 – Research Methodology 2022 – Semester 1 – Assignment 3 Question 1: Which of one of the following statements is NOT true about C-studies? C-studies research do NOT ... a) aim to explore or describe without statistical testing and typically use non-random or purposive sampling (NR). b) compare groups or examine relationships between variables by means of some form of statistical hypothesis testing. c) aim to explain phenomena by means of a hypothesis. d) use random sampling (R) as the preferred sampling technique. Question 2: Which one of the following relates to experimental designs? a) Correlational designs. b) Retrospective case study, and group-comparison. c) Quasi-experimental designs. d) Randomised cross-sectional surveys, and replicated randomised cross-sectional surveys. REMAINING ANSWERS ONLY AVAILABLE UPON PURCHASE OF THIS DOCUMENT Disclaimer: Extreme care has been used to create this document, however the contents are provided ‘as is’ without any representation or warranties, express of implied. This document is to be used for comparison, research and reference purposes ONLY. Directly submitting and/or reselling/distribution/reproduction any part of this document is not permitted. S - The study-notes marketplace Page 2 of 14 Question 3: Which one of the following statements does NOT relate to experiments in social science research? a) The researcher will do something (administer an intervention, or manipulate an independent variable, called X) to one of the groups (experimental group) and do something different (or nothing at all) to the other group (control group). b) A retrospective case study design is used whereby data is actively collected in the present. c) The groups are compared or tested for differences between them on a selected outcome, or dependent variables. This is to state that any differences in the outcome (dependent variable) between the groups are due to (or caused by) the intervention or independent variable. d) Three types of experimental designs have been developed to determine the presence of cause-and-effect relationships between different variables. The most significant differences between these types of experimental designs are with regard to: (i) degree of control on the variables being studied (X’s), (ii) degree of randomness, and (iii) presence of a comparison or control group. Question 4: The cross-sectional survey design ... a) is used exclusively in non-experimental studies. b) is an example of a standalone research design. c) is enacted as a carefully designed method for sampling the population of the study, so that the sample represents or targets specific groups of relevance in that population. d) consists of surveys questions that basically tap into the prejudices or "false-facts" that participants hold. Question 5: Threats to external validity include the following: a) History, and maturation or changes in participants. b) Demand factor, the researcher’s expectations, and the Hawthorne effect. c) Statistical regression, participant attrition, and diffusion of treatments. d) Selection, testing, and instrumentation. Question 6: The term "big data" is globally accepted for describing very large sets of information, and the application of specific computational methods for analysing such data by means of programming. Big data is described as data showing certain characteristics, widely regarded as the "Four V’s". Which one of the following statements refers to the veracity of the data? a) Data in the form of media, documents and streaming data that can be structured, unstructured or semi-structured. b) Data sets that are no longer retrospective and static but comprises live, streaming data. c) Large data sets that include pictures and video streams, such as YouTube. d) Legitimate real-time data at all times. S - The study-notes marketplace Downloaded by: creativeassignmentsic | Distribution of this document is illegal S - The study-notes marketplace Page 3 of 14 Question 7: The benefits of big data related to the integrated health and social care environment do NOT include: a) The issue of privacy and anonymisation is a highly sensitive subject within health and social care and potential for liability hinders biomedical digital research. b) Real-time data giving the provider and service user a real-time synopsis of symptoms c) Better treatment outcomes through improved diagnoses and treatment interventions d) Faster and more efficient response to the slightest emergency from the sensor or from deviations in the collected data Question 8: After establishing a research design for your study, you must identify the most suitable data collection method that fits the design and the circumstances of the research project. Which one of the following statements does NOT relate to how the data will be collected in a quantitative research approach? a) To obtain valid and reliable data you must ensure, before implementing the study, that the measurement procedures and the instruments used have acceptable levels of reliability and validity. Validity and reliability are two of the most important concepts in measurement. b) To increase the validity a researcher can increase the number of items to measure a variable. For example, by using two or more indicators (e.g. two or more questions in a questionnaire) to measure each attribute of a variable. c) Measurement implies observation of complex social phenomena by means of a numerical schema to evaluate or rate statements or items that reflect components of the phenomenon being studied. In simplified terms this means we start with some theoretical construct or phenomenon that we want to observe in human participants. This mostly abstract phenomenon is then broken down into observable indicators after which we attach number values to each indicator that will enable us to quantify our observations. d) Combinations of attributes enable us to measure the phenomenon itself. Most concepts or phenomena dealt with in the social sciences are highly abstract and not easily observable unless they are deconstructed into measurable attributes. So, if we want to measure levels of depression in human participants, we first need to deconstruct the term depression into its observable characteristics and then attach a numerical value on a scale that enables us to quantify the extent to which the characteristics are present in the participant. Question 9: The types of validity that underpin psychometric instruments or measurements are typically classified as: content, face, criterion and construct validity. Which one of the following options applies directly to construct validity? a) This means moving away from subjective assessments of face validity and providing more objective evidence of validity. This involves multiple measurements and is established by comparing scores on an instrument with an external criterion known to, or believed to, measure the concept, trait or behaviour being studied. External independent criteria are used against which to compare the scores on an instrument. If we test a new instrument for measuring a construct in people, we can also use a well-known S - The study-notes marketplace Downloaded by: creativeassignmentsic | Distribution of this document is illegal S - The study-notes marketplace Page 4 of 14 standardised measure to compare our new measure with, then we use that other measure as criterion, and this is called "concurrent validity". b) This involves determining the degree to which an instrument successfully measures a theoretical construct. Any measure exists in some theoretical context and should therefore show relationships with other constructs that can be predicted and interpreted within that context. Construct validity involves not only validation of the instrument itself, but also of the theory underlying it. To establish construct validity, the meaning of the construct must be understood and the propositions the theory makes about the relationships between this, and other constructs must be identified. One method is to compare a measure to variables it logically is NOT supposed to correlate with and to variables that we assume it should correlate with. If it works out that way, we have "convergent validity" that represents those variables the measurement strongly correlates with according to expectations. "Discriminant validity" represents variables the measurement instrument correlates poorly with. Another popular method for establishing construct validity is called "factorial validity" or "factor analysis" (DeVellis, 2017, p. 154). Factor analysis is a primary procedure to determine if a scale has internal validity based on the coherence of the underlying factors in a questionnaire. You have to understand whether underlying sets of items (latent variables) match the dimension composed into a scale. Factor analysis can either be "exploratory" or "confirmatory". The procedure determines whether items cluster together in measuring a particular construct. It also identifies items that do not relate well with the dimension of measurement. c) This is concerned with the representativeness or sampling adequacy of the content (e.g. topics or items) of an instrument. A valid measure would provide an adequate or representative sample of all content or elements or instances of the phenomenon being measured. Is the instrument really measuring the concept we assume it is? Does the instrument provide an adequate sample of items that represent the concept being measured? d) It is concerned with the superficial appearance of a measurement procedure. The relevant question in this regard is: Does the measurement technique look as if it measures the variable that it claims to measure? Question 10: A pilot test is done when we develop a draft version of the measure or use a subsection to test before implementing the study to increase reliability. This is done to... a) standardise instructions. b) standardise the conditions under which the test is taken. c) enable timely changes to increase reliability. d) increase the precision of measurement. Question 11: Which one of the following statements about reliability is INCORRECT? a) Reliability is primarily concerned not with what is being measured, but with how well it is being measured every time it is used. b) A psychological measuring instrument can be valid but not reliable. Reliability and validity exist independently from each other. S - The study-notes marketplace Downloaded by: creativeassignmentsic | Distribution of this document is illegal S - The study-notes marketplace Page 5 of 14 c) The most commonly used reliability measure using survey data is Cronbach’s coefficient alpha. This coefficient ranges between 0 and 1 and figures closer to 1 (0,8–0,9) generally indicate a highly reliable scale. d) Different reliabilities exist such as the test–retest, parallel-form, split-half, internal consistency and interrater agreement methods. Question 12: Practical considerations for collecting quantitative data include the imperative of anonymity. Anonymity is enabled by ensuring the following: a) Participation in any research should be voluntary and a respondent to a questionnaire should be able to withdraw from completing the survey at any time. b) Data collected by means of paper-based questionnaires should be stored in a lockable cabinet until such time as the data can be transferred to an electronic data set. c) There should not be any way that a participant can be identified or if there is identifiable information such as a name, or an email address, participant code or an informed consent form that may mistakenly be attached to the questionnaire, these should be separated from the questionnaire and securely stored separately. d) Electronic data should be stored on a password protected computer, not a memory stick or other storage device that is not protected. Data should only be exchanged with the analyst or statistician. On completion of the study, all research data should be archived for a prescribed period of usually five years or consistent with the organisation’s data storage policy, before it is either destroyed or deleted. Question 13: Which one of the following statements about sampling is NOT correct? a) The following steps can be followed to draw a random sample: 1. Assign a number to every person or item in the population. 2. Use a random numbers table or suitable software to select the actual sample. 3. Decide on the size of the sample; that is, what percentage of the population should be studied. 4. Identify and list the complete research population. b) By increasing sample size, you will find smaller and smaller effects to be statistically significant until, at very large sample sizes, almost any effect is significant. You must always be aware that sample size can impact on the statistical test by making it either insensitive (at small sample sizes) or overly sensitive (at very large sample sizes). Practices have been developed to statistically determine the size of a sample. Generally called power calculation, the statistical power of a study to produce sufficient effect size is calculated, either by a statistician or by using a power calculator obtainable from the internet. Sample size is influenced by different variables, of which representativeness is one. Using an insufficient number of participants results finding significant results when these are actually not significant. Using overly large samples leads to not finding significant results when these are actually present. c) The issue of the minimum size of a sample is repeatedly addressed in the literature. It is generally stated that the larger the population, the smaller the percentage of the sample from the population needs to be, and vice versa (Neuman, 2014). If the population itself is relatively small, the sample should comprise a reasonably large proportion of th