Quantitative Research by hcj


									Quantitative Research
Descriptive Research

• Also called “Survey Research”
• Describes the way things exist
• Data collected
  – Test hypothesis
  – Or to classify opinions on issues/topics
Descriptive Research

• Most common research
  – Surveys conducted
  – Many fields covered: education, politics, social
    and economic
• More complicated than one may think
Descriptive Research

• Classified by data collection method
  – Self report (survey)
  – Observation (participants watched)
Descriptive Research

• Survey types
  – Cross-sectional
     • Data collected at one point in time
  – Longitudinal
     • More than once to measure growth or change
Descriptive Research

• Cross-sectional survey
  – Single point in time may not provide
    perspective needed
  – Example: census survey
Descriptive Research

• Longitudinal survey
  – Collected multiple times from different
    samples and groups
  – Four types of longitudinal surveys
     • Trend, cohort, panel, and follow-up
Descriptive Research

• Trend survey
  – How are attitudes changing or consistent
    among changing sample groups
• Cohort survey
  – Different samples from the same group are
    surveyed over several time periods
Descriptive Research
• Panel survey
  – Same group and sample are surveyed over
• Follow-up survey
  – Similar to panel study, but performed after
    the panel study complete (examines later
  – Usually to evaluate a college instructional
Descriptive Research

• Conducting self-report research
  – Standardized, quantifiable information
  – Questionnaire is written
  – Interview is an in-person Q&A session
    between researcher and participant
Descriptive Research

• Conducting a Questionnaire study
  – Most efficient
  – Larger data sample collected
  – Mailing is preferred
Descriptive Research

• Topic under investigation must be
  significant for subjects to respond
• Problem defined in specific objectives
  – Each item on survey must directly relate to
Descriptive Research

• Participants
  – Chosen from appropriate sample
  – Possess needed information and are willing to
    share it
Descriptive Research

• Questionnaire
  – Brief, easily answered
  – Structured (close-ended) response choices
     • Facilitate easy responses
     • Data analysis objective and efficient
Descriptive Research

• Questionnaire
  – Structured items are scaled (ex. Likert)
  – Unstructured items – complete freedom of
     • Provide more depth
     • Difficult to measure
Descriptive Research

• Questionnaire
  – Item construction
     • Each questions focuses on one concept
     • Concisely worded
     • Concepts or terms with multiple connotations are
     • Avoid leading questions
Descriptive Research

• Questionnaire
  – Pretesting
     • Smaller, but similar group
     • Data will indicate any measurement difficulties and
       suggest improvement
     • Make sure it measures what it is designed to
Descriptive Research

• Questionnaire
  – Analyzing results
     • Indicate % of respondents who selected each
       alternative answer for each item
     • When possible, also analyze summed item clusters
        – Groups of items focusing on same issue
        – Variable relationships can be investigated (female vs.
          male responses) for each cluster
Descriptive Research

• Interview Study
  – More in-depth data collected if successful
  – More expensive and time-consuming
  – Smaller sample population used
Descriptive Research

• Interview Study
  – Written guide for administrator
  – Includes any prompts or additionally allowed
  – Standardized, comparable data requires all
    interviews must be conducted in same
Descriptive Research

• Interview Study
  – Uses same guidelines from questionnaire
  – Responses recorded manually or by recording
    device (participants must consent)
Correlational Research

• Data collected to determine:
  – Is there a relationship between two variables?
  – To what degree is the relationship?
• Degree of relationship is expressed as a
 correlation coefficient
Correlational Research

• Correlation coefficient
  – Measured on a scale from +1.00 to -1.00
     • If variables highly related, correlation coefficient
       near +1.00 or -1.00 will be obtained
     • If not related, a correlation coefficient near .00 will
       be obtained
Correlational Research

• Correlation coefficient
  – Interpretation
     • Lower than (+ or - ) .35 – low or none
     • Between (+ or -) .35 and .65 – moderate
     • Higher than (+ or -) .65 – High relationship
     • See table 7.1 pg. 193
Correlational Research

• Correlation coefficient
  – Interpretation
     • A high relationship between two variables doesn’t
       necessarily mean one causes the other
     • There may be a third variable that causes them
Causal-Comparative Research

• Determines cause for differences in
  behavior or status of groups of individuals
• The effect and the cause have already
• Mostly studied in retrospect
• Example: pg.217, pre-school and social
  adjustment in 1st grade
Causal-Comparative Research

• Prospective approach slightly different
  – Start with a cause and investigate its effect on
    a variable
  – Example: social promotion in 7th grade (pg.
Causal-Comparative Research

• The independent variable
  – A behavior or characteristic believed to
    influence some other behavior or
    characteristic (pre-school attendance)
• The dependent variable
  – The change or difference in a behavior or
    characteristic due to an independent variable
    (social adjustment in 1st grade)
Causal-Comparative Research

• The independent variable
  – Not manipulated in Causal-Comparative
  – This is the major difference between this and
    Experimental Research
  – In Causal-Comparative Research, groups are
    already formed and divided by the
    independent variable
Causal-Comparative Research

• The independent variable
  – Figure 8.1, pg. 219
  – Basic Causal-Comparative Research design
    finds two groups differing on some
    independent variable, and comparing them on
    a dependent variable
Causal-Comparative Research

• Groups may differ in characteristics
• Select samples that represent their
 respective populations
Causal-Comparative Research

• Data analysis and interpretation
  – Descriptive statistics
     • The mean – average performance of group on a
       measure of some variable
     • Standard deviation – measure the “spread” of
       scores around the mean (are they relatively close
       together or widely spread)
Causal-Comparative Research

• Data analysis and interpretation
  – Inferential statistics
     • t test – indicates if there is a significant difference
       between the means of two groups
     • Analysis of variance – significant differences
       between means of three or more groups
     • Chi square – does an event occur more frequently
       in one group than another?
Causal-Comparative Research

• Data analysis and interpretation
  – Use caution
  – Alleged cause-effect relationship may be the
    effect or a third factor may cause the
    independent and dependent variables
  – Correct order of causality – which variable
    caused which? The one that occurred first
Experimental Research

• Researcher manipulates at least one
  independent variable
• Controls other variables
• Observes effect on one more dependent
Experimental Research

• Independent variable
  – a.k.a. experimental variable, cause, or
  – Believed to make a difference in some
Experimental Research

• Dependent variable
  – a.k.a. criterion variable, effect, or posttest
  – Outcome of the study
  – Measure of change or difference resulting
    from independent variable manipulation
Experimental Research

• Experimental studies yield the soundest
  evidence when conducted well
• Prove or disprove hypothetical cause and
  effect relationships
Experimental Research

• The process
  – Select/define problem
  – Guided by one hypothesis
  – Researcher in control
  – Experimental group receives unique treatment
  – Control group receives different treatment
Experimental Research

• The process
  – All other variables are equal to both groups
    that may effect dependent variable
  – After exposed to treatment for some time,
    researcher applies dependent variable and
    measures differences
Experimental Research

• Manipulation and control
  – Sets experimental apart from all others
• Three forms of independent variable
  – Presence versus absence (A vs. no A)
  – Presence varying degrees (much A vs. little A)
  – Presence of one kind/another kind (A vs. B)
Experimental Research

• Manipulation and control
  – Control: efforts to remove outside influences
    of any variable (other than independent
    variable) that may affect dependent variable
Experimental Research

• Two controlled variables
  – Participant variables
     • Members of different groups may have
      characteristic differences
  – Environmental variables
     • Any change in experimental setting that may
      influence performance differences
Single-Subject Experimental
• Studies individual behavior change
• Result of some treatment form
• Repeated formula
  – Receives no treatment or intervention
  – Tested
  – Receives treatment / intervention
  – Tested

To top