Class 4: 09/19/11 building research skills (cont.)

Preview:

Citation preview

research is• We know that science evolves and

knowledge grows. The best attitude is a healthy skepticism, along with a determination to stay abreast of things.

(John Swartzberg, 2009, p. 3)

• They don't know the simple rule If it sounds too good to be true, it probably is.

(Alecia Nugent, “Too Good to BeTrue,” 2007)

all researchers must attend to

type I and type II error• type I: accepting as true what is

really false• type II: rejecting as false what is

really true

(this is the straightforward plain English version—see Vogt for the more complicated statistical version)

model: a graphic explanation of a small part of the world

• evaluating models– does this model adequately explain

new observations– does this model explain observations

better than other models

constructing a model1. hypothesize the factors involved2. hypothesize how they might be related3. draw the models4. evaluate which one explains the facts

best5. guesstimate the strengths of pathsrules• model must be falsifiable• maximize concreteness• explain as much as possible

Krathwohl: ch 7causal inference & internal integrity• associations and causation• cause:

– selecting the salient part of a causal chain for one’s inquiry

– is always inferred• Popper’s escaping disconfirmation

(falsification): no proof in empirical research

• Cronbach’s reducing uncertainty• clocklike vs cloudlike worlds

internal & external validity & cause• generalizability: based on this study

what can we say about other cases– is the relationship established in this

study– does that relationship exist elsewhere

• internal validity (internal integrity)– evidence for relationship between

variables in a study• external validity (external generality)

– whether the relationship generalizes beyond study

• operationalization: what we describe in search of the unobservable

• to establish internal validity– conceptual evidence

•explanation credibility•translation credibility

– empirical evidence•findings, results•rival explanations eliminated

– credible, logically inferred claims

validity: capable of being justified• internal

– does study do what it says it does•model theoretically sound•well thought out operationalizations•accurate descriptions (measurement or narrative)

•well thought out design•robust findings•justifiable claims•defensible conclusion

• external– do the claims generalize

•looking only at study, a judgment call

•external validity can really only be established empirically—in fact, do the claims generalize–can study be replicated in different contexts

–if we do what study claims we should, do we get the expected result

– real question not, do claims generalize, but to what

inferring causation• agreement

– what is common• difference

– what is different• concomitant variation

– do variables vary together• residuals

– after eliminating explanations, what is left

ch 8: sampling, etc

• sample and population• larger sample needed

– the greater the certainty required to infer from sample to population

– the more accurate we want to be about target population

– the more the units in the sample vary

– the smaller the effect relative to normal variation

• sampling frame—the list• sampling unit—what is selected

(unit of analysis)

probability sampling—random • stratified• systematic• cluster

– Table 8.2 useful

nonprobability sampling• judgmental• purposive• quota• snowball (chain-referral)• sequential

• the danger of convenience sampling

external validity (external generality)• (if study has no internal validity, little

point in worrying about generalizability)

• conceptual evidence– explanation generality– translation generality

• empirical evidence– “demonstrated generality”– restrictive explanations eliminated

• replicable results

Vogt• halo effect• histogram• independence• interaction effect• John Henry effect• joint probability• lurking variable• maturation effect• mortality• N! (N factorial)

Sieber: Ch 4• voluntary informed consent

– importance of gatekeepers– special populations– legal elements (see p. 33)

• effective consent statements (see p. 35)

• consent: signed, oral, or behavioral• consent as ongoing• debriefing• community-based research

test 10/03/11covers• all material in Segment I. • print lectures, take notes in class, read carefully with notes—test will

be a learning experience. • reviewing with peers, going over notes etc, will be big helpformat• available on the website Friday, 09/30 at noon (no class Monday,

10/03)• groups 2-4. turn in 3x5 index card listing group 9/26.• once you open the file, 4 hours to take the test (1 15-minute break).

do not open the test until entire group assembles • no discussion with anyone in 550 once you have begun the test until

Monday at 10 pm.• do on a computer• when finished, delete questions you chose not to answer, and print out. • put all names on test, and sign statement at the end.• hard copy in sealed 9x12 envelope (do not fold) 260 Armory by 10pm

Monday 10/03—one test per group. permitted• hand produced (by you) notes• the websitenot permitted• books• xeroxed copies of anything except 2 lists from Krathwohl ch 6

hints for test • follow directions• if 1st person singular, indicate who is speaking• (a),(b),(c) in question? then (a),(b),(c), in answer• if question about X, answer should include word

X• data plural, i.e., data are• one (total) extra point question: extra in red• be explicit in answers to questions about “your

research.” • unless question asks for answer from everyone,

one answer• “observation-N” not “subject-N”

writinglit review structure• title page• abstract • intro: repeat title (title not a heading); no heading

– describe area of interest.– specific question or problem that your review addresses– brief but detailed description of data base and strategies. State

parameters explicitly. Note possible biases from your search strategy, e.g., one type of journal.

– how review section is organized.• review section: explicit & logical scheme, e.g., sections based on topics or

types of studies. Explain. End each section with a discussion—strengths & limitations.

• discussion: Synthesize review—discussion of discussions. Communicate what you learned. Discuss general strengths and weaknesses of lit.

• conclusion: Address original question(s)—changes. Limitations of review. Implications. Areas where more or different research needs to be done

• personal reflection: Short discussion of what you have learned in the process of doing the review—about doing research, about yourself as a researcher.

Becker: preface, ch 1

• writing in grad school, and for the rest of your academic life, a very different process from writing up to this point– days of the one-draft paper are over– writing becomes a public process– writing part of the research process,

not simply what one does at the end

(Becker cont.)

• scholars know that their professional futures rests on how peers and superiors judge what they write. They can’t distance themselves from their writing.

• viewed sociologically, these writing symptoms were magical rituals—rituals to influence the result of a process over which we have no control.

• a mixed up draft is no cause for shame.• if you begin writing early in your

research, you begin to clean up your thinking earlier.

• only version that counts is the last one.

APAcitations in text • “. . . quoted text” (Chung, 2001, p. 20).• Chung (2001) noted, “Quoted text . . .”

(p. 20).

Latin abbreviations• cf. (compare), i.e. (that is), e.g. (for

example), viz. (namely), etc. (and so forth), vs. (versus) used only within parentheses. Otherwise spell out (APA, p. 108). exceptions: et al. and v. (for court cases)

citing quotation in secondary source• “Much of psychology . . . has

envisioned the child as embedded in the atemporal and acontextual realm of abstract developmental theory” (Lerner, 1998, p. 13, as cited in Lee & Walsh, 2001, p. 71).

• Lerner (1998) wrote, “Much of psychology . . . has envisioned the child as embedded in the atemporal and acontextual realm of abstract developmental theory” (p. 13, as cited in Lee & Walsh, 2001, p. 71).

• References to both Lerner (1998) and Lee & Walsh (2001) needed in References.

in American English• periods and commas always go inside

quotation marks– He said, “Please go down the hall.”

• colons and semicolons always go outside quotation marks– He wrote, “Be back soon”; then he

left.• question marks and explanation points

—place depending on the meaning– She asked, “Where are you going?”– What did she mean by “antiquated”?

use active voice• I interviewed the kids. (yes)• The kids were interviewed. (no)use first person to talk about yourself• I interviewed the kids. (yes)• The researcher interviewed the kids. (no)avoid beginning sentences with “there

is” or “it is” etc.• There were three kids who answered… (no)• Three kids answered the questions. (yes)

use who for people, that for things• I interviewed the kids, who all agreed

. . . (yes)• I interviewed the teacher that was

in . . . (no)pronouns must refer to nouns• I entered the room and found the

kids running across the table tops and throwing erasers at each other. That made me nervous. (no—not clear what made you nervous)

grad life: bestsbest cheap places to eat• Thai Eatery at the Y, Wright street

(lunch only)• Courier Café, Race St, downtown U• L’il Porgy’s Barbecue, Broadway &

University, U• Noodles, Green St., Campus Townbest video sources• That’s Rentertainment (6th & John, C)

(buy a “block”)• Urbana Free Library (downtown U), free

best coffee houses

• Caffe Paradiso, Lincoln & Nevada, U (stays open 24 hours a day during exam week)

• Café Kopi, Walnut St, downtown C

best quiet place on campus to study

• Granger Engineering library

best place for tools, kids’ clothes, and stuff in general

• Farm & Fleet, North Cunningham, U

this week: free and cheap• under construction• F: UIUC softball, 6pm

Recommended