ADVANCED RESEARCH METHODS, ANDY FIELD -
DISCOVERING STATISTICS, DISCOVERING STATISTICS
USING IBM SPSS STATISTICS, 5E, ADVANCED RESEARCH
METHODS EXAM QUESTIONS AND ANSWERS WITH
COMPLETE SOLUTIONS GRADED A++
Brivariate correlation
the relationship between two variables
relationships are best shown using:
scattergram or scatterplot
most common type of relationship only concern:
linear relationships
Positive Relationships:
two variables covary in the same direction
Negative Relationships:
two variables covary in opposite directions
Perfect Correlation:
when the points of the scattergram or scatterplot form a straight line
The statistical measures of correlation is?
Correlation Coefficients
,The Pearson's Product-moment Correlation is:
the most common measurement of linear relationship, it is a parametric linear
correlation-coefficient and has the assumption of normality
The Spearman Rank Correlation Coefficient
the non-parametric linear correlation-coefficient, used when one of both variables have
only an ordinal measurement level, can also be used when the normality assumption is
violated
The Spearman Rank Correlation Coefficient is also known as:
Spearman's Rho
What are reasons to develop a questionnaire?
to develop a survey instrument to collect a variety of data that gives you information
about a sample of people usually with a view to generalizing to a wider population
Why are surveys collected?
to obtain an estimate of the prevalence of characteristics in the population from a survey
of a representative sample AND to address research questions, test hypothesis derived
from theory
Name the 3 main components of survey design:
Questionnaire or instrument design, and sample selection, and the method of
administration (self or interviewed or internet)
Whats the first aspect of the survey needed to be decided upon?
the method of administration
Depth Interviews:
research is more qualitative in nature
,Standardized Interview:
follows a fixed protocol for the administration of a survey questionnaire
Experimental design involves:
manipulation and control of factors to allow causal conclusions
Non-experimental design involves:
measuring variables of interest and ascertain relationships
Non-experimental designs can be:
Positive, Negative, No relationship and Perfect
Non-experimental Positive designs:
variables covary together
Non-experimental Negative designs:
variables covary in opposite directions
Non-experimental no relationship designs:
variables covary independantly
Non-experimental Perfect designs:
one variable predicts another with 100% accuracy
Partial relationships:
removing the part of the relationship that is related to some other factor
What is the aim of Analysing Correlational Research?
to understand the proportion of variances in a given variable explained by other
variable/s
The most common form of a survey instrument used in psychology and social
sciences is a:
, Questionnaire
Manifest variables:
directly observable e.g. height
Latent variables:
not directly observable, must try and find indicators e.g. self-esteem
The criterion used to calculate the regression line is:
The least squares criterion
The proportion of variance in the Y variable accounted for by the X variable is
estimated by _____
rsquared
The intercept of a regression line represents _____.
the point where it crosses the Y axis when X = 0
What item difficulties is a researcher advised to aim for in constructing an ability
scale?
between .30 and .70
The concept of __________ refers to how well a test measures what it is intended
to measure, and ____________ specifically assesses whether the items on a test
measure all the knowledge or skills that are assumed to be part of the construct
measured by a test.
validity; content validity
When we refer to a test as having strong ___________, we are essentially saying
that the test has good consistency.
reliability
DISCOVERING STATISTICS, DISCOVERING STATISTICS
USING IBM SPSS STATISTICS, 5E, ADVANCED RESEARCH
METHODS EXAM QUESTIONS AND ANSWERS WITH
COMPLETE SOLUTIONS GRADED A++
Brivariate correlation
the relationship between two variables
relationships are best shown using:
scattergram or scatterplot
most common type of relationship only concern:
linear relationships
Positive Relationships:
two variables covary in the same direction
Negative Relationships:
two variables covary in opposite directions
Perfect Correlation:
when the points of the scattergram or scatterplot form a straight line
The statistical measures of correlation is?
Correlation Coefficients
,The Pearson's Product-moment Correlation is:
the most common measurement of linear relationship, it is a parametric linear
correlation-coefficient and has the assumption of normality
The Spearman Rank Correlation Coefficient
the non-parametric linear correlation-coefficient, used when one of both variables have
only an ordinal measurement level, can also be used when the normality assumption is
violated
The Spearman Rank Correlation Coefficient is also known as:
Spearman's Rho
What are reasons to develop a questionnaire?
to develop a survey instrument to collect a variety of data that gives you information
about a sample of people usually with a view to generalizing to a wider population
Why are surveys collected?
to obtain an estimate of the prevalence of characteristics in the population from a survey
of a representative sample AND to address research questions, test hypothesis derived
from theory
Name the 3 main components of survey design:
Questionnaire or instrument design, and sample selection, and the method of
administration (self or interviewed or internet)
Whats the first aspect of the survey needed to be decided upon?
the method of administration
Depth Interviews:
research is more qualitative in nature
,Standardized Interview:
follows a fixed protocol for the administration of a survey questionnaire
Experimental design involves:
manipulation and control of factors to allow causal conclusions
Non-experimental design involves:
measuring variables of interest and ascertain relationships
Non-experimental designs can be:
Positive, Negative, No relationship and Perfect
Non-experimental Positive designs:
variables covary together
Non-experimental Negative designs:
variables covary in opposite directions
Non-experimental no relationship designs:
variables covary independantly
Non-experimental Perfect designs:
one variable predicts another with 100% accuracy
Partial relationships:
removing the part of the relationship that is related to some other factor
What is the aim of Analysing Correlational Research?
to understand the proportion of variances in a given variable explained by other
variable/s
The most common form of a survey instrument used in psychology and social
sciences is a:
, Questionnaire
Manifest variables:
directly observable e.g. height
Latent variables:
not directly observable, must try and find indicators e.g. self-esteem
The criterion used to calculate the regression line is:
The least squares criterion
The proportion of variance in the Y variable accounted for by the X variable is
estimated by _____
rsquared
The intercept of a regression line represents _____.
the point where it crosses the Y axis when X = 0
What item difficulties is a researcher advised to aim for in constructing an ability
scale?
between .30 and .70
The concept of __________ refers to how well a test measures what it is intended
to measure, and ____________ specifically assesses whether the items on a test
measure all the knowledge or skills that are assumed to be part of the construct
measured by a test.
validity; content validity
When we refer to a test as having strong ___________, we are essentially saying
that the test has good consistency.
reliability
