Chapter 2 Research Methods Miller Intimate Relationships 6e

  • Slides: 37
Download presentation
Chapter 2 Research Methods Miller Intimate Relationships, 6/e Mc. Graw-Hill/Irwin Copyright (c) 2012 by

Chapter 2 Research Methods Miller Intimate Relationships, 6/e Mc. Graw-Hill/Irwin Copyright (c) 2012 by The Mc. Graw-Hill Companies, Inc. All rights reserved.

Relationship Research There is a lot of personal opinion — rather than scientific fact

Relationship Research There is a lot of personal opinion — rather than scientific fact — competing for your attention. • Relationship science can affect us personally in ways most other sciences do not. • Be Better Consumers of Science: Discriminating the viable and shoddy 2 -2

A Brief History of Relationship Science The scientific study of human relationships is a

A Brief History of Relationship Science The scientific study of human relationships is a relatively new endeavor, beginning in earnest only in the 1960 s, but has made tremendous strides in the last 45 years. For that reason, relationship science is less well-known, and perhaps less appreciated, than most other sciences. 2 -3

A Brief History of Relationship Science Nevertheless, family scholars, psychologists, sociologists, and communication researchers

A Brief History of Relationship Science Nevertheless, family scholars, psychologists, sociologists, and communication researchers now examine varied types of intimate relationships: • often using diverse samples of people from all walks of life • often studying them over long time periods • studying both pleasant and unpleasant aspects • often following them in their natural settings 2 -4

A Brief History of Relationship Science Look in your library for the journals: Journal

A Brief History of Relationship Science Look in your library for the journals: Journal of Marriage and the Family Journal of Social and Personal Relationships And visit the International Association for Relationship Research at: http: //www. iarr. org 2 -5

Developing a Question The questions researchers ask emerge from various sources: • • Personal

Developing a Question The questions researchers ask emerge from various sources: • • Personal experience Social problems Previous research Theories 2 -6

Developing a Question Some studies seek to describe events as they naturally occur… And

Developing a Question Some studies seek to describe events as they naturally occur… And other studies strive to establish the causal connections between events that determine which causes have what consequences. Ø A thoughtful consumer judges an investigation with regard to its intended purpose. 2 -7

Obtaining Participants Whose relationships get studied? Convenience Sample: anyone who is readily available –

Obtaining Participants Whose relationships get studied? Convenience Sample: anyone who is readily available – the oft common sample Representative Sample: a group of people about whom we want to make conclusions. 2 -8

Obtaining Participants A common problem in relationship research: Volunteer Bias: Of those invited, people

Obtaining Participants A common problem in relationship research: Volunteer Bias: Of those invited, people who agree to participate may differ from those who refuse. Example: Karney et al. (1995) found that volunteers were better educated, employed in higher-status jobs, and more likely to have cohabited, than those who refused to participate in a longitudinal study of their relationships 2 -9

Obtaining Participants Because representative samples are expensive, convenience samples are used more often. BUT….

Obtaining Participants Because representative samples are expensive, convenience samples are used more often. BUT…. Is this a big problem for relationship science? 2 -10

Obtaining Participants The rationale used by researcher who use volunteer subjects like college students:

Obtaining Participants The rationale used by researcher who use volunteer subjects like college students: Ø Many processes studied by relationship scientists are so basic that they don’t differ much from group to group. 2 -11

Obtaining Participants An example from 1989: Who said “yes”? yes Men Women “Would you

Obtaining Participants An example from 1989: Who said “yes”? yes Men Women “Would you go out with me tonight? ” “Would you come over to my apartment tonight? ” “Would you go to bed with me tonight? ” 50% 69% 56% 6% 75% 0% (Clark & Hatfield, 1989) 2 -12

Obtaining Participants What should we make of one study at just one university that

Obtaining Participants What should we make of one study at just one university that only included college students? Caution is warranted. (But by the way, other investigations involving more than 20, 000 people from every major region of the world have since gotten very similar results. ) 2 -13

Obtaining Participants Ø As long as we are thoughtful and judicious, even small studies

Obtaining Participants Ø As long as we are thoughtful and judicious, even small studies using convenience samples can make valuable contributions to our science. Ø No one study is perfect. Ø Be cautious. Ø Diverse methods and samples are valuable. Science is cumulative. Each small step is helpful. 2 -14

Research Development • Constructs : No agreement on how to measure abstract concepts (e.

Research Development • Constructs : No agreement on how to measure abstract concepts (e. g. , anxiety, love, attraction, etc. ) – Operational Definition: declaration of how the construct will be measured • Example: Measure Anxiety: Measure heart rate, Galvanic skin response, self report, observation 2 -15

Choosing a Design Several different research designs are used by relationship science: • Correlational

Choosing a Design Several different research designs are used by relationship science: • Correlational designs • Experimental designs • Developmental designs – Cross-sectional designs – Longitudinal designs – Retrospective designs 2 -16

Choosing a Design Correlational designs measure naturallyoccurring events, looking for associations between them. •

Choosing a Design Correlational designs measure naturallyoccurring events, looking for associations between them. • Relationship between two dependent variables • In a positive correlation, increases in one event are associated with increases in the other. • In a negative correlation, increases in one event are associated with decreases in the other. 2 -17

Choosing a Design A correlation tells us that two events are associated, but it

Choosing a Design A correlation tells us that two events are associated, but it does not explain causal connections, if any, between them. Why are two events related? A simple correlation can’t tell us why. - Power of Predictability : one variable is strongly associated with the other - No Causal Power: Third Variable Problem (e. g. , bars and churches) 2 -18

Choosing a Design • Experimental designs manipulate a variable to delineate clearly the effect

Choosing a Design • Experimental designs manipulate a variable to delineate clearly the effect it causes • The variable manipulated is the independent variable • They illuminate cause and effect. “If we change x , what happens to y ? ” 2 -19

Choosing a Design Experiments can be very informative, but experimenters must be able to

Choosing a Design Experiments can be very informative, but experimenters must be able to control and manipulate the events they wish to study. You can’t do experiments on events you cannot control… …such as how much two people love each other. 2 -20

Choosing a Design Developmental designs study the manner in which events change over time.

Choosing a Design Developmental designs study the manner in which events change over time. • A cross-sectional design compares different groups of people who are at different stages or ages in a developmental process. • A longitudinal design studies the same group of people with repeated measurements over a period of time. • A retrospective design asks people about their past experiences. 2 -21

Choosing a Design Developmental Designs: Each of these designs has advantages and disadvantages. Ø

Choosing a Design Developmental Designs: Each of these designs has advantages and disadvantages. Ø No one type of research design is always perfect. 2 -22

Selecting a Setting Laboratories allow researchers control over events, but may elicit unnatural behavior.

Selecting a Setting Laboratories allow researchers control over events, but may elicit unnatural behavior. Natural environments, such as a couple’s home, may elicit more typical behavior but be full of distractions. 2 -23

Selecting a Setting Some structured situations invite people to role-play (i. e. , to

Selecting a Setting Some structured situations invite people to role-play (i. e. , to pretend certain events are happening). • In scenarios, people read a story and imagine it taking place. • In simulations, people act out a particular role. • In immersive virtual environments, people interact with 3 -D computer representations of other people. 2 -24

The Nature of the Data What data are actually being collected? • • •

The Nature of the Data What data are actually being collected? • • • Self-reports Observations Couples’ reports Physiological measures Archival evidence 2 -25

The Nature of the Data Self-reports: Asking people about their feelings and behavior •

The Nature of the Data Self-reports: Asking people about their feelings and behavior • Self-reports are inexpensive and easy to obtain • Self-reports help us understand people’s personal points of view, giving invaluable information to help us understand the workings of relationships 2 -26

The Nature of the Data Potential problems with self-reports: • Participants may misunderstand some

The Nature of the Data Potential problems with self-reports: • Participants may misunderstand some questions. • Their memories may not be accurate. • People may not be able or willing to tell the truth. 2 -27

The Nature of the Data Potential problems with self-reports: Self-serving bias: Participants may honestly

The Nature of the Data Potential problems with self-reports: Self-serving bias: Participants may honestly think that they are kinder, more competent, and more attentive than they really are. Social desirability bias: People may be reluctant to report anything that makes them look bad. 2 -28

The Nature of the Data Observations: watching behavior directly Scientific observations are not casual

The Nature of the Data Observations: watching behavior directly Scientific observations are not casual undertakings. Observers are highly trained, and their watching is often highly detailed. 2 -29

The Nature of the Data Observations Experience-sampling uses short, intermittent periods of observation to

The Nature of the Data Observations Experience-sampling uses short, intermittent periods of observation to capture samples of behavior that occur over longer periods of time. 2 -30

The Nature of the Data Observations Reactivity is sometimes a problem: People may change

The Nature of the Data Observations Reactivity is sometimes a problem: People may change their behavior when they know they are being observed. 2 -31

The Nature of the Data Physiological Measures assess partners’ autonomic and biochemical responses. Measures

The Nature of the Data Physiological Measures assess partners’ autonomic and biochemical responses. Measures of heart rate, stress hormones, or genital arousal can provide information about how interactions with others affect people physically. 2 -32

The Nature of the Data Archival Materials Inspection of personal documents (such as yearbooks

The Nature of the Data Archival Materials Inspection of personal documents (such as yearbooks and diaries) and public records (such as marriage licenses) can also be informative. 2 -33

The Nature of the Data Couples’ Reports involve self-reports of one’s own behavior and

The Nature of the Data Couples’ Reports involve self-reports of one’s own behavior and observations of the other’s behavior from both members of a couple. Sometimes there is little agreement between the partners’ reports (e. g. , he might say he did something that she says did not occur). 2 -34

The Ethics Should relationship science pry into people’s personal affairs? Relationship scientists are very

The Ethics Should relationship science pry into people’s personal affairs? Relationship scientists are very careful to protect the welfare of the participants in their research… …but people may be asked about sensitive matters. 2 -35

The Ethics of Such Endeavors Arguably, relationship science has an ethical imperative to gain

The Ethics of Such Endeavors Arguably, relationship science has an ethical imperative to gain knowledge that can benefit us. Too many close relationships fail, so it would be unethical not to try to understand how good relationships work. 2 -36

Research Design Task Group Assignment: See Handout Turn in one sheet per group with

Research Design Task Group Assignment: See Handout Turn in one sheet per group with group members names. 2 -37