Developing biodata questionnaires Developing biodata questionnaires
4. Developing a pool of biodata items◦ From previous biodata research studies or originally
developed
a) Habits and attitudes
b) Health
c) Human relations
d) Money
e) Parental home, childhood, teenage years
f) Personal attributes
g) Present home, spouse, and children
Developing biodata questionnaires Developing biodata questionnaires
4. Developing a pool of biodata itemsh) Recreation, hobbies, and interests
i) School and education
j) Self impressions
k) Values, opinions, and preferences
l) Work
◦ Journal of applied psychology
◦ Personnel psychology
Developing biodata questionnaires Developing biodata questionnaires
Guidelines 1. Should principally deal with past behavior and
experiences
2. Family relationships are usually viewed as offensive
3. Specificity and brevity
4. Numbers should be used to define a biodata item’s options or alternatives
5. All possible response options or an “escape” option (other) should given; response options that form a continuum should be used.
Developing biodata questionnaires Developing biodata questionnaires
Guidelines 6. Neutral or pleasant connotation
7. Past and present behaviors and with opinions, attitudes, and values are generally acceptable
8. Reflect historical events that are important in shaping a person’s behavior and identity
9. To lessens the effect of socially desirable
Developing biodata questionnaires Developing biodata questionnaires
5. Prescreening and pilot-testing biodata items◦ Reviewed by a panel of judges◦ Large sample sizes ( >500)◦ 刪題標準a) Little response variance
b) Skewed response distributions
c) Items correlated with protected-group characteristics such as ethnicity
d) Having no correlation with other items
e) No correlation with the criterion
Scoring biodata questionnairesScoring biodata questionnaires
Computing a single biodata score◦ As empirical keying◦ Methods: (a) vertical percentage (b) horizontal
percentage (c) correlation (d) differential regression (e) deviant response (f) rare response
◦ vertical percentage - item is analyzed related to job success weight sum weight
◦ Limited to predictions for a specific pool of applicants and for a specific job success
◦ Uncertainty about what the scores really means.
Scoring biodata questionnairesScoring biodata questionnaires
Computing a multiple biodata score◦ 用主成份分析或因素分析 ( 通常為 10~15 因素 )◦ 每個因素包含至少 10 題以上◦ Figure 10.4 p455◦ 有學者分析 75 biodata items 和三個效標的關係◦ Different biodata factors were associated with different
types of performance on the job.
信度信度
因為個別差異, internal consistency reliability 0.6 ~ 0.8
Test-retest reliability 0.6 ~ 0.9
Objective, verifiable biodata factors had higher test – retest reliabilities than subjective , less verifiable ones.
效度效度
Edwin Ghiselli 認為是最有效的預測工具James Asher reported that 90 percent of biodata
validity coefficient were above 0.3Meta-analyses Different biodata validity coefficients for
different criteria can be expected Table10.3 p458The predictive validity of biodata measures was
0.35 for overall job performance and 0.30 for overall performance in training programs
效度效度
效度類化,有學者使用 supervisory profile record 在 79 個公司,效度為 0.33
應同一公司經理人較恰當, true validity 0.53
Guidelines◦ 效度類化合理嗎 ?◦ 效標的測量穩定及準確有效嗎 ?◦ 每一題的效度 ?◦ 樣本數夠大嗎 ?
效度效度
以上的研究都是指在職經理人,可以類化到應徵者嗎 ?
Biodata measure 和 general mental ability measure correlation 0.5 , biodata 並無顯著增加預測力
但工作表現廣泛定義包含組織公民行為,這是 biodata 可以預測到的
Accuracy of biodataAccuracy of biodata
Self – report
Attitudes, interests, belief – not verifiable
但也有學者支持有正確性
Greater accuracy was found for scores on groups of common biodata items – biodata factor – than for single item
Accuracy of biodataAccuracy of biodata
應徵者會基於 impression management strategies fake answer
避免策略◦ Score specific biodata item response◦ 告訴應徵者會在面試時驗證◦ Use of factual items – more verification ◦ p461
Why is biodata a good predictorWhy is biodata a good predictor
BIB 反應人的 life history, interview 容易扭曲
Item is validity
Measure aspects of the criterion directly
Composed of a wide variety of items.
Use of biodataUse of biodata
0.4% of 248 HR
Table 10.4 p465
Criticism of biodataCriticism of biodata
Fake
計分方式都 specific, biodata form 未統一
只有大公司才有大樣本
有些人才會被排除
Summy recommendation for WABsSummy recommendation for WABs
Plan on developing your own WAB
Consider carefully the measure of job performance or criterion to be used in developing a WAB
Develop and select WAB items with the content of the job and criterion to be predicted in mind.
Summy recommendation for WABsSummy recommendation for WABs
Develop a WAB on as large as sample possible ( a minimum of 300 application forms from persons hired and on whom criterion data are available )
Avoid using the same sample of persons for whom a WAB was developed to cross-validity a WAB scoring key.
Summy recommendation for WABsSummy recommendation for WABs
If your organization is using expensive selection measures – for example, assessment centers – consider using a WAB as a preliminary screening tool.
Use WABs in conjunction with other selection measures.
Conduct a validity check of WAB scoring keys every three to five years.
Summy recommendation for Summy recommendation for biodatabiodataConduct a job analysis; consider carefully the
criterion or job success measure to be predicted.
Consider using an exiting biodata questionnaire.
However, because few well-developed, commercially available biodata questionnaires are currently available, biodata questionnaires will probably have to be developed,
Development of biodata questionnaires is more suited to larger organizaions.
Summy recommendation for Summy recommendation for biodatabiodataItems or constructs developed for a biodata
questionnaire must be related empirically to the criterion.
An important problem facing small organizations is how best to examine of the validity.
Like a WAB, all biodata items should be reviewed for discriminatory impact.
Summy recommendation for Summy recommendation for biodatabiodataIn conducting a validation study, base the study
on job applicants, use predictive validity designs, and cross-validate the questionnaire on applicants
Consider using biodata questionnaire in addition to other measures; biodata questionnaire may explain unique criterion information unexplained by other measures, such as interviews and certain ability tests.