Subjective measures of social status - harry ganzeboom's home page

Subjective measures of socialstatus: validationHarry B.G. GanzeboomFree University AmsterdamISSP, Taipei, April 29 2007Three measures of subjectivesocial status• Centers (1949): Subjective 'Social Class (lowerclass, working class, upper working class, middleclass, upper middle class, upper class).• Diagram question: "Where do you belong indiagram?" (1 – top … 6 – bottom).• Top-bottom ladder: "Where would you putyourself on this scale?" (1 – top … 10 – bottom).Subjective social status in ISSP 1Subjective social status in ISSP 2Questions• Do they measure the same thing (validity)?• Do they measure that same thing with equalreliability?• Do we need three indicators: what is thatgood for?Research design• Compare distributions across country andtime:– R2 by country– Change in means over time• Missing value patterns• Reliability / factoi analysis• Validity analysis in structural modelSubjective social status in ISSP 3Subjective social status in ISSP 4EducationStructural modelMeasurement modelSUBJECTIVESTATUSOccupationSUBJECTIVESTATUSYIncomeclass diafit topbotSubjective social status in ISSP 5Subjective social status in ISSP 6

Data (from INEQ modules)Missing values analysis• Distributional analysis: all available data.• Validation: only countries/wave in which allthree measures are available–Pooled– Per country/wave19871992CLASS2.910.6DIAFIT16.7TOPBOT3.63.219994.56.83.3Subjective social status in ISSP 7Subjective social status in ISSP 8Correlations – all dataCorrelations -- listwiseCorrelationsCorrelationsTOPBOTCLASSDIAFIT1Pearson CorrelationSig. (2-tailed)NPearson CorrelationSig. (2-tailed)NPearson CorrelationSig. (2-tailed)NTOPBOT CLASS DIAFIT11 .487 .651. .000 .00070613 56537 37996.487 1 .508.000 . .00056537 58141 36925.651 .508 1.000 .000 .37996 36925 38880TOPBOTCLASSDIAFIT1Pearson CorrelationSig. (2-tailed)NPearson CorrelationSig. (2-tailed)NPearson CorrelationSig. (2-tailed)NTOPBOT CLASS DIAFIT11 .511 .650. .000 .00036145 36145 36145.511 1 .510.000 . .00036145 36145 36145.650 .510 1.000 .000 .36145 36145 36145Subjective social status in ISSP 9Subjective social status in ISSP 10Reliability• Alpha: 0.767• Without CLASS: 0.744• Without DIAFIT1: 0.675• Without TOPBOT: 0.630Differences and trends• TOPBOT 18.7% wave: ns• DIAFIT1 14.2% wave: t = -2.1• CLASS 13.6% wave: t = +2.5Subjective social status in ISSP 11Subjective social status in ISSP 12

Predicting TOLARGECorrelations with objective status[1] [2]• CLASS .139 .079• DIAFIT1 .144 .104• TOPBOT .089 .072• SUBJPOS .214 .214CLASSDIAFIT1EDUC-ATION.231.161OCCUP-ATION.226.158INCOME.145.198(Standardized regression coefficients, controlled for countryand wave differences, [1]: bivariate, [2] multiple.)TOPBOT.130.141.184Subjective social status in ISSP 13Subjective social status in ISSP 14Intervening variable SES TOLARGE?Total effectControllingCLASSControllingDIAFIT1ControllingTOPBOTControllingSUBJPOS.101.046.061.062.034.149.181.167.218Observe• Effect of subjective position is considerablystronger than of objective position.• Three indicators are needed to bring out theintervening effect.Subjective social status in ISSP 15Subjective social status in ISSP 16