Soc004 Exam 1

Selective Observation

Only noticing things that are in line with own preferences/beliefs

inaccurate observation

An observation based on faulty perceptions

Overgeneralization

Unjustifiably concluding that what is true is true for most/all cases

Illogical reasoning

drawing conclusions from invalid assumptions

Resistance to change

the reluctance to change our ideas in light of new information

Scientific process

1.review literature

2.identify research question

3. Derive hypothesis from theory (for deductive research)
4. Collect own data / use secondary data ( experiment, survey, observation, analysis, interview)
5. Analyze data

6.Draw Conclusions

7.Share results

Purpose of research

exploration, description, explanation(does x cause y?)

Four Norms of science

1. Communalism

2. Organized Skepticism

3. Disinterestedness

4. Universalism

Universalism

• 1 of 4 norms of science

• acceptance/rejection of scientific claims should not depend on personal or social characteristics of those making the claim

Universalism Example

Dr.Berry concluded that a study he recently read must be a high quality study because the study author works at a prestigious university.

• Conclusion inconsistent w/ universalism b/c this should not depend on the author’s academic background. Just because he went to a prestigious university does not mean he is correct

Communalism

• 2 of 4 norms of science

• Scientific findings should be the collective property of the entire scientific community

Disinterestedness

• 3 of 4 norms of science

• pursuit of scientific truth should proceed without personal bias or motivation

Organized skepticism

• 4 of 4 norms of science

• claims to truth should be approached from position of doubt

Theory

a logically interrelated set of propositions about empirical reality

Hypothesis

Tentative statement about empirical reality involving a relationship between two or more variables

Variables

characteristic or property that can vary (i.e., take on different values/attributes)

Attributes

• characteristics; qualities of a person or thing

• Attributes should be mutually exclusive (no overlapping attributes) and exhaustive (every case should fit into a category)

Example of Exhaustive/Exclusive

Highest educational degree obtained?

• no degree

• certificate

• HS Diploma

• BA

• PhD

• other

-Exhaustive b/c every case would fit in one of these categories

-Exclusive b/c it’s asking for HIGHEST degree so every one should fit in only one category

Example of non exclusive variable

Ex. region of residence: rural, urban, north, and south

→ not mutually exclusive ( one can be in multiple categories ex. Urban & south)

Example of non Exhaustive variable

EX. Highest educational degree obtained: PhD and BA

→ not exhaustive B/c some may only have HS diploma/ no degree

-categories will not fit every case

Exhaustive Variables example

every case should fit into a category

• Religious affiliation:

  • Incorrect - Christian ;Buddhist; Muslim; Jewish

  • Correct - Christian; Buddhist; Muslim; Jewish ; no affiliation; other

Example of Mutually Exclusive variable

No overlapping attributes

• Incorrect: 0-20, 20- 40, 40 - 60, 60 +

• Correct: 0-20, 21-40, 41-60, 61+

Independent Variable (IV)

variable that is hypothesized to cause a dependent variable

Ex. Poverty Rate

Dependent Variable (DV)

The outcome variable; variable that is hypothesized to depend on or be caused by the IV

Ex. percentage on community residents who are homeless

Finding IV & DV in a Hypothesis

rephrase to an if - then statement : "If the IV increases ( or decreases), then the Dv increases (or decreases)"

IV/DV if-then example 1

The greater the use of internet, the greater the strength of distant family ties (positive relationship; increase - increase)

• IV - level of internet use

• DV: Strength of distant family ties

• If-Then - If internet use greater, then strength of distant family ties greater

IV/DV if-then example 2

Risk of property theft decreases as income increases (Negative relationship decrease - increase)

• IV: Income

• DV: Risk of property theft

• If-then - if income is higher, then risk of property theft is less

positive relationship

-variables move in same direction

-Increase in one variable causes increase in another

-Decrease in one variable causes decrease in another

Positive relationship example 1

the more you practice , the better you get (more practice→more skill)

Positive relationship example 2

The less hours you work, the less your paycheck will be( less hours→less money)

Negative relationship

-Variables move in opposite direction

-Increase in one variable causes decrease in another

-Decrease in one variable causes increase in another

Negative relationship example 1

The more you smoke; the unhealthier you are ( more smoke ; less healthy)

Negative relationship example 2

The less you smoke , the healthier you are ( smoke down ; health up)

Deductive research

• a hypothesis is derived from a theory & then tested

• Begins with theory → directs study

• tests theory

inductive research

-after data is collected, a researcher (induces) a general explanation ( a theory) to account for the data

-Starts with data --> data used to develop theory

-Builds theory

Institutional Review Board (IRB)

• Oversees treatment of human subjects; reviews research proposals that involve human subjects

• a group of organized & community representatives required by federal law to review the ethical issues in all proposed research on human subjects conducted in institutions that receive federal funds.

Ethical guidelines for research

1. Achievement of valid result

2. Ethics of data collection, analysis,& reporting
   - honest, open: procedures (eventually) data
   -(scientific norm of communalism)
   -pursuit of knowledge above personal gain/promotion of ideology
   - (scientific norm of disinterestedness)

3. Ethical treatment of human subjects
   - overseen by institutional review board (IRB)

Guidelines on the ethical treatment of human subjects

1.Avoid harming participants
2. Obtain information consent
3.Avoid deception (except in limited circumstances)
4. Maintain privacy

1. Avoid harming participants

-Difficult to define
-Hard to predict
-Cost-benefit analysis (how will/is study useful?)
-Protect participants from harm by:

• Inform of risk through inform consent process

• Screen out vulnerable participants

• Asses in debriefing; offer follow-up counseling

2. Obtain informed consent

• Participants voluntarily decide to participate based on full understanding of possible risks involved

• Scientific cost to full disclosure?/Dealing with deception study?

• Do not provide full disclosure but debrief

3. Avoid deception

-Sometimes part of experimental design

Disclose during debrief

-Cost-benefit analysis

4. Maintain privacy/confidentiality

-Anonymity(participants anonymous) or confidentiality (not anonymous but promise not to publicly disclose info) unless voluntarily & explicitly waived

• limits

  • laws may allow records to be subpoenaed & may require reporting child abuse, health/life threatening situations

  • exception - certificate of confidentiality

Certificate of Confidentiality

a certificate issued to a researcher by the NIH that ensures the right to protect information obtained about high-risk populations or behaviors—except child abuse or neglect—from legal subpoenas.

ASA Code of Ethics

1. Professional competence

2. Integrity

3. Professional and scientific responsibility

4. Respect for people's rights, dignity, and diversity

5. Social responsibility

6. Human Rights

Conceptualization

• process of specifying what we mean by a term

• Concept —> definition

Concept

•  mental image that summarizes a set of similar observations, feelings, or ideas, indicators, overlapping dimensions

Operationalization

• Specifying the measure that will indicate the value of cases on a variable

• Definition —> operational Indicators

  • Selection

    • questions

    • coding schemes for content analysis

    • Coding scheme for observations

    • Unobtrusive measure

    • Manipulations (experiments) - researchers manipulate IV (predictor)/ researcher manipulate experiments

EX: Conceptualization \Operationalization: Internal locus of Control

• internal locus of control

  • Conceptualization —> feeling control over outcomes in one’s life

  • Operationalization: “Agree/Disagree to following statements”

    • There’s no sense planning a lot– if something good is going to happen, it will - Internals disagree

    • I am responsible for my own success - Internals agree

    • I have little control over the bad things that happen to me - Internals disagree

    • I can do anything I really set my mind to - Internals agree

EX. Conceptualization/Operationalization: Trust

• Trust

  • Conceptualization - Believing most people are good & unlikely to harm other people

  • Operationalization - “Please rate level of agreement w/ each statement using following scale

    1———-2——-3——-4——-5——-6——-7——8——-9—-10

    Strongly agree Agree Neutral Disagree Strongly Disagree

Concept, Variables, and Indictors example

• Concept - Binge drinking

• Variable - frequency of heavy episodic drinking

• Indicators - “how many times have you had 5+ drinks in a row during past 2 weeks”

Levels of Measurement

1. Nominal

2. Ordinal

3. Interval

4. Ratio

5. Dichotomies (dummy variables)

Nominal Measures

variable attributes only have characteristics of exhaustiveness & exclusiveness (categorical variables) ; attributes only named - weakest

• No mathematical interpretation -Don’t rank order, vary in kind/quality but not amount

  • gender (female, male,..)

  • Race (African, Hispanic, ...)

  • Religion ( Christian, Hindu, Buddhist,..)

Nominal example

 Region of Birth in U.S.  - Northeastern, Midwest, South, Southwest, West, other)

• categories , no rank order, “other” can be used as key word

Ordinal Measure

Variable whose attributes can be somewhat rank order

• SES (Upper, Middle, Working, ..)

• Education (phd, BA, HS Diploma,..)

• Agreement w/ attitudinal statements ( strongly agree, agree, disagree, strongly disagree)

Ordinal Examples

• What is your political orientation - very liberal, somewhat liberal, somewhat conservative, very conservative

• How wrong do you think it is for someone your age to take a handgun to school?  - Very wrong , Wrong, a little bit wrong, Not wrong at all

Interval Measure

• variable whose atrributes are rank ordered & have equal distances between adjacent attributes ; meaningful distance

  • EX - temperature , ( 50, 75, 100,..) - 0 temp. Does not mean no temperature, just value in scale 

  • IQ score

  • Dates

Interval Example

• Example 1 - When did you start your current position? _____ / _____/___

Month        Day   Year

Ratio Measurement

• variable whose attributes are rank-order, equal distances between adjacent attributes, are based on a “true zero” point

• EX: 

  • Age

  • Income

  • Frequency measures

  • # of siblings

  • Weekly Hrs of TV

Dichotomies

• (Dummy Variables) variable having only two values, coded as 0 and 1

  • Treated as ratio

    • Example

      • Employed = 1 ; not employed = 0

      • Democrat = 1 ; not democrat = 0

        • Example 1: Are you a veteran?

          • Yes____1

          • No___2

Measurement Reliability

• consistency of measure; degree to which the same data would have been collected each time in repeated observations of the same phenomenon

  • Are participants understanding questions?

  • Is majority understanding?

• Reliable measure does not have to be valid

Measurement Validity

congruence between concepts and measures; it exists when a measure measures what we think it measures

• In other words, if we seek to describe the frequency of domestic violence in families, we need to develop a valid procedure for measuring domestic violence.

• Valid Measure MUST be reliable

Measurement reliability/validity example

• Ex. prof uses tape measure to measure knowledge of methods course

  • Tape measure reliable in measuring height - will measure height accurately constantly  

  • Not valid bc measuring height has nothing to do with sociology/course

Checking for Measurement consistency

• 1.Across multiple times

  • Test-retest

• 2.Across multiple indicators

  • Inter-item

• 3. Across multiple observers

  • Interobserver-intercoder

1. Multiple times

• Test-retest reliability

  • strong positive correlation between both scores on a measure administrated to same individuals at two points in time

    • Termed intra-rater/intra-observer reliability when applied to raters/observers

  • Problem - Concept must not change or else reliability score will be lower than it should be as score from test 1 different from scores from test 2

    • Participants may remember/mindlessly repeat answers making reliability score higher than it should be

  • Fix Problesm

    • Alternate-forms ; Adjust timing

2. Multiple Indicators

Inter-item reliability

• strong correlation among all items in composite measure (used to measure single concept)

  • Suggest homogenous set of items

3. Multiple observers

Intercoder/Interobserver reliability

• strong correlation in data coding results/scores produced by different observers/coders

types of Validity Assessments

1. Face Validity

2. Content Validity

3. Construct Validity

4. Convergent Validity

1. Face validity

Inspection of items used to measure concept suggest that they are valid (almost Intuitively) 

• When measure is obviously pertains to the meaning of concept being measured

  • Has face validity b/c on face of item/surface it seems valid 

2. Content Validity

 Inspection of items used to measure concept suggest that they cover full range of the concept’s meaning

• EX- measure knowledge of class at end of quarter put only have concept of measure on test

  • Low content validity b/c not tested on other topics in lectures like variables, research, experiments, ethics, etc

  • High content validity if  tested on various topics/ all topics covered throughout quarter

3. Construct Validity

scores on a measure are correlated with other variables in a theoretically expected manner.

• Correlations with related variables

  • Just world scale

    • negative correlation

  • Scale measuring of prejudice toward African Americans

    • negative correlation

4. Convergent Validity

Measure is related to different measure of same concept

• Ex. Use 2 measures of depression that show similar results / ; measure same concept

  • Use new measure developed and use existing depression measure

  • If new measure shows depression than old measure should show same/similar results