Docstoc

p4

Document Sample
p4 Powered By Docstoc
					METODE DE CERCETARE
      • CURS 4
•CORELATIE si CAUZALITATE


    DESIGN EXPERIMENTAL


•CONCEPTE, OPERATIONALIZARI, VARIABILE
                     Ipoteze de cercetare

I. sint exprimari explicite cu privire la modul in care cercetatorul
crede ca fenomenele de interes sint legate intre ele;
- presupuneri despre modul in care variabilele independente le
influenteaza pe cele dependente.
Exprima modul in care variatia variabilei dependente poate fi
explicata prin cea a variabilei independente.
Forma generala este:

H:     A il determina pe B (intr-un anumit fel)
Caracteristici ale ipotezelor bune
1. Au un caracter empiric si nu normativ: sint despre ce este si
nu despre ce ar trebui sa fie.

2. generalitate: afirmatii nu despre un senator, dar despre
senatori in general, nu despre o tara dar despre tari in
general, nu despre un razboi ....

3. plauzibilitate

4. consistente cu datele

5. sint testabile

6. este precizata relatia dintre variabile: sens, forma,
intensitate.
       COUNTRY:               9.00 Norway                                                                                                                             COUNTRY:        8.00 Slovenia
1.2                                                                                                                                                             1.2



1.0                                                                                                                                                             1.0




                                                                                       Involvement typology (recoded 0-1)
                                                                                                                                                                 .8
 .8

                                                                                                                                                                 .6
 .6

                                                                                                                                                                 .4
 .4

                                                                                                                                                                 .2
 .2

                                                                                                                                                                0.0
0.0
                                                                                                                                                                -.2
-.2                                                                                                                                                               0.0         20.0        40.0        60.0   80.0    100.0
  10.0       20.0      30.0     40.0     50.0                                   60.0                                        70.0          80.0   90.0
                                                                                                                                                                      AGE
       AGE


       COUNTRY:               10.00 Romania                                                                                                                     COUNTRY:             7.00 Netherlands
1.2                                                                                                                                                       1.2


1.0                                                                                                                                                       1.0
                                           Involvement typology (recoded 0-1)




 .8                                                                                                                                                        .8


 .6                                                                                                                                                        .6


 .4                                                                                                                                                        .4


 .2                                                                                                                                                        .2


0.0                                                                                                                                                       0.0

-.2
                                                                                                                                                          -.2
      0.0           20.0          40.0                                          60.0                                               80.0           100.0
                                                                                                                                                            0.0             20.0         40.0         60.0    80.0     100.0

       AGE                                                                                                                                                      AGE
H: Democratic characteristics are influenced mostly by
durable cultural traditions.

These traditions are associated with several attributes that can be
employed in order to classify nations or sub-national communities:
(1) the dominant religion (Huntington 1996),
(2) the empire in which ancestors of the present citizens lived for
centuries (Reisinger 1999, and Lipset 2000),
(3) the length of time spent under a Leninist regime (Fuchs 1999b and
Rohrschneider 1999),
(4) the level of socio-economic modernity (Lipset 1959, 1994, 2000),
and
(5) the level of civicness in the past (Putnam 1993).


http://www.policy.hu/badescu/paper1.rtf
Differences between Transylvania and the rest of the country with respect of past
attributes:
1. Transylvania was part of the Habsburg Empire until 1918, whereas the rest of the
country belonged to the Ottoman Empire until 1877, and remained independent after.
2. Transylvanians have been of higher heterogeneity in terms of ethnicity and religious
denominations, compared with the rest of the country. In 1930, 57.8% of
Transylvanians were Romanians, compared with 88.5% in the rest of the country. In
1857 about half of the Romanians from Transylvania were Greek-Catholics and the
other half were Orthodox, most of the other Transylvanians being Catholics and
Protestants, whereas more than 90% of the Romanians from the rest of the country were
orthodox. The communists disbanded the Greek Catholic Church in late 1948, and
forced it to merge with the Orthodox Church.
3. Transylvania entered to the twentieth century with a significantly higher level of
socio-economic modernity, compared with the rest of the country:
          a. Demographical facts which tend to be associated with levels of socio-
          economic modernity.
          b. The literacy rate was higher in Transylvania than in the rest of the country:
          51.1% compared to 39.3%, in 1897-1912, and 67% compared to 55.8%, in
          1930 (Livezeanu 1998, p.48).
          c. The degree of associativness, and perhaps that of civicness, seemed to have
          been higher in Transylvania that in the rest of the country.
The effects of region and size of locality on the proportion of the Romanians who
voted in 1996 general elections for PDSR, and CDR respectively, in two multilinear
regression models (the unit of analysis is locality).

                           PDSR (%)                 CDR (%)
                           b       beta    S.E.     b        beta     S.E.
Region                     -20.01 -0.70 0.37        6.50     0.33     0.34
Size of locality           -0.12 -0.14 0.01         0.12     0.19     0.01
Region: people from Transylvania are coded with '1', those from the other region with '0'.
Size of locality: the number of Romanians above 18 years (thousands).
Citizen support of a set of criteria for a democratic community.
                       DEM            AUT        PSC         CGI       VIO               LAW
Romania                88.6           22.0       11.0        15.8      77.0              77.8
Region 1               85.1           24.2       12.7        14.9      74.6              78.3
Region 2               90.3           19.2       10.0        16.3      78.7              77.7
DEM: Support of democratic rule; AUT: Support of autocratic rule; PSC: Support of
political system of one's own country; CGI: Confidence in governmental institutions;
VIO: Illegitimacy of violence; LAW: Law abidingness.

Citizen support of different types of democratic community at cultural level.
                       SRE          SOL          TRU            WET        ETO
Romania                31           63           18             63         6
Region 1               29.8         52.7         19.6           60.5       6.7
Region 2               28.7         67.4         19.3           64.6       6
SRE: Self-responsibility; SOL: Solidarity; TRU: Trust in others; WET: work ethic; ETO:
ethic tolerance.

Citizen support for different types of democratic community at structural and
process level.
                     PRO         MAN         PMO          CIV
Romania              55          37          21           9
Region 1             56.3        33.1        18.1         11.1
Region 2             55.2        37.9        20.3         3.8
PRO: private ownership; MAN: management of enterprise; PMO: political motivation;
CIV: civic engagement.
The effect of region on political values and behavior in 1998 data, in
multilinear regression models.
Variable                                  Beta        t adj. R2
Democrat 1: v157                           0.16   4.48      0.04
Democrat 2: v163                           0.07   1.96      0.01
Autocrat 1: v154                          -0.04 -1.20       0.05
Autocrat 2: v156                          -0.16 -4.55       0.04
Support for political system: v152         0.04   1.10      0.01
Confidence in government: inst            -0.06 -1.83       0.08
Illegitimacy of violence: v164             0.00   0.04      0.02
Law abidingness: law                       0.02   0.60      0.03
Self-responsibility: v127                 -0.04 -1.34       0.02
Solidarity: v172                          -0.09 -2.61       0.02
Work ethic: v129                           0.02   0.66      0.01
Ethic tolerance: tolerant                 -0.11 -3.41       0.10
Private: v126                             -0.06 -1.74       0.02
Management: v87 *                          0.07   1.90      0.01
Political motivation: polit                0.04   1.10      0.04

The effect of region on political values and behavior in 1993 and 1998 WVS data, in
logistic regression models.
                                                 1993                       1998
                                           b      S.E.     exp.b      b      S.E.     exp.b
Civic engagement                           0.08     0.20      1.08     1.02    0.15     2.77
Trust: v27                                -0.23     0.18      0.79     0.02    0.17     1.02
Petitions                                                              0.57    0.19     1.77
Boycotts                                                               0.62   -0.39     1.86
Demonstrations                                                         0.32    0.16     1.37
H: A este o cauza a lui B.

Este comparat un grup de obiecte caracterizat de o
anumita valoare a lui A cu un alt grup de obiecte, cat mai
asemanator cu primul, caracterizat de o alta valoare a lui
A.

Ipoteza este sustinuta daca B se manifesta diferit intre
cele doua grupuri de obiecte, in sensul prezis de
ipoteza.
Experimentul
____________________________________________
                    T1          X          T2
grup de control                           


grup experimental                         
____________________________________________
T1: pretestare
T2: testare
X: factorul care actioneaza asupra grupului experimental dar nu si asupra celui de
control

Ex: imaginati un design experimental prin care sa poata fi testata afirmatia:
Schumaher este cel mai bun pilot de F1.
Exemplu:

Efectul informarii asupra preferintelor electorale in
  alegerile locale.
(ex. Stanley Milgram,
http://www.new-life.net/milgram.htm)



Today the field of psychology would deem this study highly
unethical but, it revealed some extremely important findings.
The theory that only the most severe monsters on the sadistic
fringe of society would submit to such cruelty is disclaimed.
Findings show that, "two-thirds of this studies participants fall
into the category of ‘obedient' subjects, and that they represent
ordinary people drawn from the working, managerial, and
professional classes (Obedience to Authority)."
Ultimately 65% of all of the "teachers" punished the "learners"
to the maximum 450 volts. No subject stopped before reaching
300 volts!
Ex. Simulare “Crime and Punishment”
Imaginati o cercetare (ipoteza, design
experimental), care sa foloseasca aceasta
simulare.
Probleme care apar cand sunt realizate experimente:

1. E dificil sa fie selectate grupuri asemanatoare
2. Actiunea asupra grupului experimental este imposibila
   sau dificila
3. Situatia reprodusa prin experiment este artificiala
4. Probleme de etica
Concepte, operationalizari, variabile

concept - un simbol al unui obiect sau a unei proprietati a unui obiect,
sau a unui fenomen comportamental. (Nachmias & Nachmias)

Conceptele contribuie la delimitarea unei discipline:
fizica: masa, gravitatie, forta psihologie: personalitate, devianta sociologie: clasa
sociala, mobilitate demografie: mortalitate, fertilitate stiintele politice: democratie,
putere, reprezentare, justitie, egalitate.
granitele nu sint rigide: ex. socializare, comportament rational, calitatea vietii,
saracie

Aceiasi termeni pot sa desemneze categorii diferite in discipline diferite: haos
(matematica/stiinte sociale), capital social (economie/stiintele politice).

cele mai multe concepte din s.p. sint imprecise iar intelesurile care li se atribuie pot
sa difere intre politologi: democratie, liberalism, drepturile omului, integrare politica.
de aceea astfel de concepte trebuie definite, precizate de fiecare data cind sint
folosite.
definitie conceptuala: definitie care descrie conceptele facind apel
la alte concepte.

definitie operationala: descriere a unei multimi de proceduri pe
care cercetatorul trebuie sa le urmeze pentru a stabili existenta
fenomenului descris de concept.

ex. inteligenta    - capacitatea de adaptare rapida la situatii noi
                   - performanta la un test de inteligenta
                   (set de intrebari+modalitatea de interpretare a raspunsurilor).

Exemple:
A. participare politica: acele activitati ale cetatenilor particulari care sint
intreprinse cu scopul de a influenta selectia personalului guvernului sau a
actiunilor intreprinse de acesta (Verba si Nie, Participation in America)
Operationalizare: Verba si Nie, pag.351
Exemplu:
Ne intrebam daca tarile tind sa fie mai democratice atunci cind sint mai dezvoltate
economic. Avem la dispozitie date referitoare la 12 tari (notate a, b, c, d, ... l). Atit
pentru notiunea de "tara democratica" cit si pentru cea de "tara dezvoltata
economic" avem la dispozitie cite doua definitii:

tara democratica
def1. multipartitism, alegeri libere, participare publica in procesul de guvernare.
def2. libertatea cuvintului, a presei, religiei, etc.

tara dezvoltata economic
def 1. proportia ridicata populatiei angajate in sectorul tertiar
def 2. produsul national brut pe cap de locuitor

Q. Interpretati rezultatele din perspectiva problemei de studiat, pentru fiecare din
cele 4 variante.
                                         ECONOMIE
                                              DEF1                  DEF2
DEMOCRATIE                            dezv.      nedezv.    dezv.      nedezv.
                 DEF1       dezv.    abcdef                 abc            def
                           nedezv.               ghijkl     ghi            jkl
                 DEF2       dezv.      def           ghi    ghi            def
                           nedezv.    abc            jkl    abc            jkl




  Q. Interpretati rezultatele din perspectiva problemei de studiat, pentru
  fiecare din cele 4 variante.
VARIABILE: estimari (masuri) ale unor proprietati empirice .

Ex: sex, virsta, proprietatea de a raspunde intr-un anumit fel la o intrebare (da, foarte
mult...), numarul de cuvinte dintr-un articol, mortalitatea infantila a unei tari, numar de
conflicte armate intr-o regiune intr-o perioada, rata de participare la vot pe sectii de
votare.

variabila, cazuri, valori. Variabila se refera la o multime de obiecte (oameni, raspunsuri
ale unor oameni, tari, conflicte armate) - pe care le numim cazuri - si la un atribut
comun tuturor aceste obiecte. Pentru fiecare caz variabila ia o valoare.

Ex. 1000 de persoane sint intrebate:
in ce masura sinteti de acord cu afirmatia: locul femeii este la bucatarie.
1. complet de acord 2. oarecum de acord 3.nu prea sint de acord 4. deloc de acord.

variabila independenta: masuri ale unor fenomene despre care se crede (se presupune)
ca influenteaza, afecteaza, cauzeaza alte fenomene.
variabila dependenta: var. despre care se incearca sa se arate ca este influentata de
var.ind.

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:32
posted:9/19/2012
language:Latin
pages:22