Page |1
PYC3714 Psychological Research
TOPIC 1
STUDY GUIDE ACTIVITIES AND SUMMARY
Summary of important points in this topic
After studying this topic, you should be familiar with the following terms and ideas, and how they are used in psychological
research:
❖ Constructs: concepts that act as explanations for phenomena, events and behaviour and are abstracted from
observations.
❖ Theories: a theory is a frame of reference for facts that attempts to account for why things are as they are; a claim
about how constructs are related to produce phenomena, which has been validated by research.
❖ Measurement: allocating numbers according to a rule in order to quantify constructs. In order to measure
something, a way must be found to operationalise it.
❖ Variables: constructs that were measured and are represented by numbers; the variable makes the latent
construct manifest to the researcher. There can also be hidden variables of which the researcher may be unaware.
❖ Measurement error: a consequence of the fact that our measurements are imperfect, and we use samples of data
to represent populations. The variable we observe contains an unknown amount of measurement error.
❖ Data: collections of information, which could be quantities representing many measurements for many individuals
or objects.
❖ Descriptive statistics vs. inferential statistics: using certain values to summarise data and making use of
statistical techniques to make inferences or generalisations about them.
❖ Populations and samples: the whole group to which a set of measurements refers can be represented by using a
smaller group to represent them.
❖ A research hypothesis: a statement about a possible relationship among constructs that may explain some set of
observations that one intends to investigate.
In each of the following statements of a research problem, identify which one is the dependent and which one the independent variable:
i. Couples who are willing to openly discuss their feelings, have more stable marriages.
ii. Men are more likely to be aggressive than women.
iii. People who were abused as children are more likely to abuse their own children than people who were not abused as
children.
Answers:
i. Here, `willingness to discussing feelings' is supposed to have an effect on `stability of marriage'. Therefore `stability of
marriage' is the dependent variable and `willingness to discussing feelings' is the independent variable.
ii. In this statement, the claim is made that `expression of aggression' is affected by `gender', so gender is the independent
variable on which the expression of aggression (the dependent variable) depends.
iii. In this statement the claim is made that the predisposition to abuse one's 8 children (the dependent variable) is dependent
on whether one was abused as a child (the independent variable).
Imagine two groups of school children are compared on the construct `mathematical ability' using an appropriate standardised test,
one group coming from an up-market suburb in a big city, and the other group from a rural environment. Let us assume that, using the
techniques of statistical hypothesis testing from this course, the researcher finds a difference exists in favour of the urban sample.
Can you think of any hidden variable - factors that may influence the children's mathematical ability - beyond the fact that they come
from either an urban or a rural environment? Consider it for a moment, before reading our comment below.
Comment: What the researcher should consider is that a hidden variable may exist: the group of urban school children may have
access to better quality education. Note that the difference that the statistical test shows could well be real, but the conclusion could
be false, because the hidden variable `quality of mathematics tuition' is not considered. It is not a false relationship; it is just that we
cannot assume that the one variable affects the other one directly based purely on the fact that we found that relationship. Such hidden
variables are sometimes also referred to as nuisance variables, because they interfere with the ability of the researcher to make
sensible conclusions.
, Page |2
Keep in mind that techniques of statistical inference can show you that a difference (or other interaction effect) exists, but it cannot tell
you why it exists. The reasoning is up to you. The procedures do not replace the need for critical thinking.
Let’s closely look at variables using examples:
Find the cause and effect: IV / DV?
• The higher you go, the more the temperature increases
• The more hours you study, the higher your marks in the exam
• The taller you are, the more intelligent you
Other examples
• The more time you spend in the sun, the more likely you’ll get skin cancer
• The darker you hair, the healthier its roots
• Why are there/why are there not any extraneous/confounding variables in the above?
Answers:
Find the cause and effect: IV / DV?
• The higher you go, the more the temperature increases: IV (altitude), DV (temperature)
• The more hours you study, the higher your marks in the exam: IV (hours of study), DV (exam marks)
• The taller you are, the more intelligent you: IV (height), DV (intelligence)
Other examples
• The more time you spend in the sun, the more likely you’ll get skin cancer: IV (time/hours in the sun), DV (likelihood of skin
cancer)
• The darker you hair, the healthier its roots: IV (colour of hair), DV (health of roots)
• Why are there/why are there not any extraneous/confounding variables in the above? Confounding variables need to be
identified and stated before any experiment is conducted. For example, in the former examples, a family history of skin cancer
has not been ruled out as a contributing variable. While in the latter example, daily use of shampoo or other hair treatment
products has also not been considered as a confounding variable.
Look at the table with data related to the AIDS scenario in Appendix A. This is typical of how data would be presented in a spreadsheet.
You cannot really make much sense of it just by looking at it. Just looking at one or more pages of numbers tells you very little. So the
problem is how to make sense of it, how to use it to form conclusions about the underlying constructs that were measured to produce
these numbers, and the possible patterns that exist among them.
Let us suppose you want to do research on work stress of single mothers. You want to know whether their stress levels have an
influence on how strictly they discipline their children. Write down a description that defines the appropriate population in your study.
Then look at our answer below. Also see if you can identify the dependent and independent variables in this study.
Answer
The population would be all women who are not currently married, who have children and who have a job. The fact that they have
stress (or not) is not part of the definition of the population, since stress level is one of the variables you want to study. You should not
only look at single working mothers who are stressed, but rather at single working mothers who have a range of levels of stress. Stress
level is in fact the independent variable, with strictness of discipline as the dependent variable. This is because you want to know
whether increased stress affects discipline, not so much the other way round.
An example of a research hypothesis would be: `motorists are more likely to express frustration at delays on the road on work days
than on weekends'. Which are the constructs that may be at work here?
It looks as though there are two clear constructs: the amount of frustration that motorists show, and a dichotomous variable (see
Appendix B if you are not sure what this means) dividing the time they are observed into `work days' and `weekends'.
An educational psychologist has the idea that students (at a residential university) who review their work every day after the relevant
lectures will do better in exams than students who only look at their notes a few days before the exam, trying to study everything in
one intense burst of concentration. This is based on the psychologist's own observations, but also seems to be in line with a theory
on optimal memory strategies that she supports.
, Page |3
Try to consider this problem on your own. First, write out the research question as you understand it, then think about which
constructs and variables are relevant. Then try to write out the research question that is implied as a research hypothesis. (We suggest
you write out your thoughts on a separate piece of paper before you look at our attempt to answer it below.)
Answer:
Here, the general statement that forms the research question could be: How do those students who review their work regularly after
the relevant lectures do in their exams relative to the students who do not do so, but prefer to study all their material in one intense
burst of concentration? Does the first group do better than the second group in their exams, as the memory theory seems to suggest?
Multiple-choice questions
1. Inferential statistics is a branch of statistics concerned with ...
1. inferring numerical properties of sample data.
2. inferring properties of samples from assumptions.
3. estimating properties of populations from data.
2. Research in psychology is primarily about ...
1. gathering facts.
2. testing theories of human behaviour.
3. making assumptions.
3. The main purpose of psychological research is to ...
1. test theories empirically.
2. select random samples.
3. apply inferential statistics.
4. A theory is a network of ...
1. relations among facts that were proven to be true.
2. explanations for observed phenomena in terms of constructs.
3. hypotheses that were observed.
5. Constructs are ...
1. theoretical in nature.
2. empirical in nature.
3. identical to variables.
6. An item in a psychological test may be considered to be a ...
1. manifest variable.
2. test construct.
3. both of the above.
7. Constructs such as anxiety can be ...
1. observed directly.
2. empirical in nature.
3. defined in terms of behaviour.
8. Operational definitions of psychological constructs should define
constructs ...
1. in terms of observable behaviour.
2. in terms of other constructs.
3. through measurement.
9. Operational definitions enable us to ...
1. bridge the gap between theory and observations.
2. observe constructs.
3. do both of the above.
10. The nature of an hypothesis refers to a rule that tells us ...
1. which variables cause, or are associated with, which variable.
2. which value of an independent variable is associated with which value of a dependent variable.
, Page |4
3. the nature of the variables as constructs.
11. Suppose an hypothesis states that X causes Y. We cannot predict Y exactly because ...
1. of unknown population or sample values.
2. of other variables that also influence Y.
3. we might not know the values of X.
Consider the following hypothesis and then answer items 12 and 13: Hypothesis: The viewing of violent video material is related to
aggressive behaviour.
12. The independent variable in the hypothesis above is ...
1. viewing of violent video material.
2. aggressive behaviour.
3. There is no independent variable.
13. The hypothesis above ...
1. associates values of the independent variable with values of the dependent variable.
2. implies which levels of `aggressive behaviour' are caused by viewing violent video material.
3. is a theory about how violence on videos and aggressive behaviour are connected.
14. The observation that `a child hits another child' can be considered to be ...
1. an observation of a manifest variable.
2. a behavioural consequence of `aggressive behaviour'
3. both of the above.
15. An hypothesis is general or universal in nature in the sense that ...
1. the postulated relation is linear.
2. the relation is proposed for a population.
3. the postulated relation will hold in a specific sample.
16. A sample is ...
1. a segment of a population
2. a representation of the population
3. random
17. An inference is ...
1. another term for hypothesis
2. an inspired guess
3. a generalisation based on existing information
18. `Operationalisation' refers to ...
1. the process of forming an hypothesis
2. the process of finding a practical way of measuring a construct
3. finding a practical operation or procedure to do the research
19. To say that a construct is `latent' is another way of saying it is ...
1. hidden from direct observation
2. abstract
3. a concept that forms part of a theory
20. Parameters are ...
1. another word for descriptive statistics
2. values that indicate certain important aspects of the data obtained from a sample
3. values that summarise aspects of population data
21. In psychological research, a theory is best understood as ...
1. a list of constructs
2. an educated guess
3. an explanation of how facts are connected
22. To operationalise a construct means to ...
PYC3714 Psychological Research
TOPIC 1
STUDY GUIDE ACTIVITIES AND SUMMARY
Summary of important points in this topic
After studying this topic, you should be familiar with the following terms and ideas, and how they are used in psychological
research:
❖ Constructs: concepts that act as explanations for phenomena, events and behaviour and are abstracted from
observations.
❖ Theories: a theory is a frame of reference for facts that attempts to account for why things are as they are; a claim
about how constructs are related to produce phenomena, which has been validated by research.
❖ Measurement: allocating numbers according to a rule in order to quantify constructs. In order to measure
something, a way must be found to operationalise it.
❖ Variables: constructs that were measured and are represented by numbers; the variable makes the latent
construct manifest to the researcher. There can also be hidden variables of which the researcher may be unaware.
❖ Measurement error: a consequence of the fact that our measurements are imperfect, and we use samples of data
to represent populations. The variable we observe contains an unknown amount of measurement error.
❖ Data: collections of information, which could be quantities representing many measurements for many individuals
or objects.
❖ Descriptive statistics vs. inferential statistics: using certain values to summarise data and making use of
statistical techniques to make inferences or generalisations about them.
❖ Populations and samples: the whole group to which a set of measurements refers can be represented by using a
smaller group to represent them.
❖ A research hypothesis: a statement about a possible relationship among constructs that may explain some set of
observations that one intends to investigate.
In each of the following statements of a research problem, identify which one is the dependent and which one the independent variable:
i. Couples who are willing to openly discuss their feelings, have more stable marriages.
ii. Men are more likely to be aggressive than women.
iii. People who were abused as children are more likely to abuse their own children than people who were not abused as
children.
Answers:
i. Here, `willingness to discussing feelings' is supposed to have an effect on `stability of marriage'. Therefore `stability of
marriage' is the dependent variable and `willingness to discussing feelings' is the independent variable.
ii. In this statement, the claim is made that `expression of aggression' is affected by `gender', so gender is the independent
variable on which the expression of aggression (the dependent variable) depends.
iii. In this statement the claim is made that the predisposition to abuse one's 8 children (the dependent variable) is dependent
on whether one was abused as a child (the independent variable).
Imagine two groups of school children are compared on the construct `mathematical ability' using an appropriate standardised test,
one group coming from an up-market suburb in a big city, and the other group from a rural environment. Let us assume that, using the
techniques of statistical hypothesis testing from this course, the researcher finds a difference exists in favour of the urban sample.
Can you think of any hidden variable - factors that may influence the children's mathematical ability - beyond the fact that they come
from either an urban or a rural environment? Consider it for a moment, before reading our comment below.
Comment: What the researcher should consider is that a hidden variable may exist: the group of urban school children may have
access to better quality education. Note that the difference that the statistical test shows could well be real, but the conclusion could
be false, because the hidden variable `quality of mathematics tuition' is not considered. It is not a false relationship; it is just that we
cannot assume that the one variable affects the other one directly based purely on the fact that we found that relationship. Such hidden
variables are sometimes also referred to as nuisance variables, because they interfere with the ability of the researcher to make
sensible conclusions.
, Page |2
Keep in mind that techniques of statistical inference can show you that a difference (or other interaction effect) exists, but it cannot tell
you why it exists. The reasoning is up to you. The procedures do not replace the need for critical thinking.
Let’s closely look at variables using examples:
Find the cause and effect: IV / DV?
• The higher you go, the more the temperature increases
• The more hours you study, the higher your marks in the exam
• The taller you are, the more intelligent you
Other examples
• The more time you spend in the sun, the more likely you’ll get skin cancer
• The darker you hair, the healthier its roots
• Why are there/why are there not any extraneous/confounding variables in the above?
Answers:
Find the cause and effect: IV / DV?
• The higher you go, the more the temperature increases: IV (altitude), DV (temperature)
• The more hours you study, the higher your marks in the exam: IV (hours of study), DV (exam marks)
• The taller you are, the more intelligent you: IV (height), DV (intelligence)
Other examples
• The more time you spend in the sun, the more likely you’ll get skin cancer: IV (time/hours in the sun), DV (likelihood of skin
cancer)
• The darker you hair, the healthier its roots: IV (colour of hair), DV (health of roots)
• Why are there/why are there not any extraneous/confounding variables in the above? Confounding variables need to be
identified and stated before any experiment is conducted. For example, in the former examples, a family history of skin cancer
has not been ruled out as a contributing variable. While in the latter example, daily use of shampoo or other hair treatment
products has also not been considered as a confounding variable.
Look at the table with data related to the AIDS scenario in Appendix A. This is typical of how data would be presented in a spreadsheet.
You cannot really make much sense of it just by looking at it. Just looking at one or more pages of numbers tells you very little. So the
problem is how to make sense of it, how to use it to form conclusions about the underlying constructs that were measured to produce
these numbers, and the possible patterns that exist among them.
Let us suppose you want to do research on work stress of single mothers. You want to know whether their stress levels have an
influence on how strictly they discipline their children. Write down a description that defines the appropriate population in your study.
Then look at our answer below. Also see if you can identify the dependent and independent variables in this study.
Answer
The population would be all women who are not currently married, who have children and who have a job. The fact that they have
stress (or not) is not part of the definition of the population, since stress level is one of the variables you want to study. You should not
only look at single working mothers who are stressed, but rather at single working mothers who have a range of levels of stress. Stress
level is in fact the independent variable, with strictness of discipline as the dependent variable. This is because you want to know
whether increased stress affects discipline, not so much the other way round.
An example of a research hypothesis would be: `motorists are more likely to express frustration at delays on the road on work days
than on weekends'. Which are the constructs that may be at work here?
It looks as though there are two clear constructs: the amount of frustration that motorists show, and a dichotomous variable (see
Appendix B if you are not sure what this means) dividing the time they are observed into `work days' and `weekends'.
An educational psychologist has the idea that students (at a residential university) who review their work every day after the relevant
lectures will do better in exams than students who only look at their notes a few days before the exam, trying to study everything in
one intense burst of concentration. This is based on the psychologist's own observations, but also seems to be in line with a theory
on optimal memory strategies that she supports.
, Page |3
Try to consider this problem on your own. First, write out the research question as you understand it, then think about which
constructs and variables are relevant. Then try to write out the research question that is implied as a research hypothesis. (We suggest
you write out your thoughts on a separate piece of paper before you look at our attempt to answer it below.)
Answer:
Here, the general statement that forms the research question could be: How do those students who review their work regularly after
the relevant lectures do in their exams relative to the students who do not do so, but prefer to study all their material in one intense
burst of concentration? Does the first group do better than the second group in their exams, as the memory theory seems to suggest?
Multiple-choice questions
1. Inferential statistics is a branch of statistics concerned with ...
1. inferring numerical properties of sample data.
2. inferring properties of samples from assumptions.
3. estimating properties of populations from data.
2. Research in psychology is primarily about ...
1. gathering facts.
2. testing theories of human behaviour.
3. making assumptions.
3. The main purpose of psychological research is to ...
1. test theories empirically.
2. select random samples.
3. apply inferential statistics.
4. A theory is a network of ...
1. relations among facts that were proven to be true.
2. explanations for observed phenomena in terms of constructs.
3. hypotheses that were observed.
5. Constructs are ...
1. theoretical in nature.
2. empirical in nature.
3. identical to variables.
6. An item in a psychological test may be considered to be a ...
1. manifest variable.
2. test construct.
3. both of the above.
7. Constructs such as anxiety can be ...
1. observed directly.
2. empirical in nature.
3. defined in terms of behaviour.
8. Operational definitions of psychological constructs should define
constructs ...
1. in terms of observable behaviour.
2. in terms of other constructs.
3. through measurement.
9. Operational definitions enable us to ...
1. bridge the gap between theory and observations.
2. observe constructs.
3. do both of the above.
10. The nature of an hypothesis refers to a rule that tells us ...
1. which variables cause, or are associated with, which variable.
2. which value of an independent variable is associated with which value of a dependent variable.
, Page |4
3. the nature of the variables as constructs.
11. Suppose an hypothesis states that X causes Y. We cannot predict Y exactly because ...
1. of unknown population or sample values.
2. of other variables that also influence Y.
3. we might not know the values of X.
Consider the following hypothesis and then answer items 12 and 13: Hypothesis: The viewing of violent video material is related to
aggressive behaviour.
12. The independent variable in the hypothesis above is ...
1. viewing of violent video material.
2. aggressive behaviour.
3. There is no independent variable.
13. The hypothesis above ...
1. associates values of the independent variable with values of the dependent variable.
2. implies which levels of `aggressive behaviour' are caused by viewing violent video material.
3. is a theory about how violence on videos and aggressive behaviour are connected.
14. The observation that `a child hits another child' can be considered to be ...
1. an observation of a manifest variable.
2. a behavioural consequence of `aggressive behaviour'
3. both of the above.
15. An hypothesis is general or universal in nature in the sense that ...
1. the postulated relation is linear.
2. the relation is proposed for a population.
3. the postulated relation will hold in a specific sample.
16. A sample is ...
1. a segment of a population
2. a representation of the population
3. random
17. An inference is ...
1. another term for hypothesis
2. an inspired guess
3. a generalisation based on existing information
18. `Operationalisation' refers to ...
1. the process of forming an hypothesis
2. the process of finding a practical way of measuring a construct
3. finding a practical operation or procedure to do the research
19. To say that a construct is `latent' is another way of saying it is ...
1. hidden from direct observation
2. abstract
3. a concept that forms part of a theory
20. Parameters are ...
1. another word for descriptive statistics
2. values that indicate certain important aspects of the data obtained from a sample
3. values that summarise aspects of population data
21. In psychological research, a theory is best understood as ...
1. a list of constructs
2. an educated guess
3. an explanation of how facts are connected
22. To operationalise a construct means to ...
