Research Methods and Design Questions and Answers

Answer: B

Limitations laid down by the researcher are called delimitations.

Delimitations are the boundaries a researcher deliberately sets on a study: which population, which variables, which area and which period will be covered. They are choices made in advance to keep the study manageable. (Limitations, by contrast, are weaknesses outside the researcher's control.)

Objectives state what the study intends to achieve. Assumptions are things accepted as true without testing them. A hypothesis is a tentative statement of the expected result. None of these sets the boundaries of the study.

Answer: B

The technique is the survey.

A survey collects primary data directly from a sample of respondents, usually through a questionnaire or interview schedule, about their characteristics, opinions or behaviour.

Observation records behaviour by watching it rather than asking questions, simulation imitates a real situation in a model, and experimentation manipulates variables under control to test cause and effect.

Answer: C

A cohort study design is longitudinal.

In a cohort study a group of people who share a characteristic (for example, born in the same year or exposed to the same factor) is followed over time, and outcomes are recorded at intervals. Following the same group over time is what makes it longitudinal.

It is observational, not experimental, since the researcher does not assign exposures. Although it describes and explores, its defining feature among these options is the follow-up over time.

Answer: B

If a measure has high validity, it must have high reliability.

Reliability is consistency of measurement and validity is measuring what is intended. An unreliable measure gives inconsistent results, so it cannot measure the intended thing accurately. Reliability is therefore a necessary condition for validity: every valid measure is reliable, though a reliable measure need not be valid.

The other three options all say reliability is unimportant or optional, which contradicts this relationship.

Answer: D

The highest level of measurement is the ratio scale.

Stevens's four levels are nominal (categories), ordinal (order), interval (equal distances) and ratio (equal distances and a true zero). The ratio scale has all the properties of the lower ones plus an absolute zero, so ratios such as 'twice as heavy' are meaningful; weight, height and income are ratio data.

Nominal and ordinal are the lowest levels, and the 'internal scale' (meaning interval) lacks a true zero, so it ranks below ratio.

Answer: A

Thick and thin data refer to the qualitative and quantitative approach.

'Thick' data, a term drawn from Clifford Geertz's 'thick description', are rich qualitative data about context, meanings and emotions, from interviews or ethnography. 'Thin' data are numerical, quantitative data, such as large datasets that show what happens but not why.

Options B and C each name only one side of the pair, and 'soft data' covers only the qualitative half.

Answer: A

An observation schedule is a set of explicit rules for assigning behaviour to categories.

In structured observation the researcher uses a schedule that defines the categories of behaviour to be recorded and the rules for deciding which category each observed act belongs to, so that different observers code the same behaviour in the same way.

It is not a timetable of observation days, and a list of questions for an interviewee is an interview schedule. Testing measurement validity is a separate step that may use the data, not the schedule itself.

Answer: B

The key's characteristic of action research is that data is qualitative.

Action research is carried out by practitioners, such as teachers or extension workers, to solve a problem in their own setting. It typically relies on qualitative data such as observations, reflective journals, interviews and discussions, gathered in a cycle of plan, act, observe and reflect.

Controlled variables and proof of cause and effect belong to experimental research, and because action research concerns one particular local situation, its results are not meant to be generalised.

Answer: B

Studying a phenomenon over different periods of time is longitudinal methodology.

A longitudinal study observes the same subjects or phenomenon repeatedly over an extended period, which shows change, development and trends, as in studies of child growth.

Experimental research manipulates variables to test cause and effect, analytical research critically examines existing facts, and explanatory research seeks reasons for a relationship. None of them is defined by repeated study over time.

Answer: D

The appropriate method is stratified random sampling.

The study compares the prevalence of diabetes among males and females, so the population should be divided into strata by sex and a random sample drawn from each stratum. This ensures both groups are adequately represented in each survey year and allows a valid comparison.

Cluster sampling selects whole groups such as villages and does not guarantee balance by sex. Systematic sampling takes every kth unit from a list, and convenience sampling uses whoever is available, which is biased and unsuitable for estimating prevalence.