psych 475 ex2 lickin

What is measurement in psychology?

The assigning of numbers to objects or events. Can be both categorical or numerical.

What is a construct and indicator and what is an example of it?

Measures the concept of something like job performance.

Meanwhile, the second term is the question that measures the construct. For example, “do you like your job?”

What is nominal measurement?

Assigning identification symbol, no order nor math

ex: for gender,

1= female

2= male

What is ordinal measurement?

Assigns numbers to stuff with natural order. Cannot use math because there is no equal spacing

Ex: bachelors<masters<doctorate<

What is interval measurement?

Allows for addition and subtraction, “how much” comparisons. However, no division/multiplication

Ex: temperature, IQ scores

You can say someone’s 20 IQ points smarter, not 2x smarter

What is ratio measurement?

All properties of interval measurement and includes division and multiplication due to the zero point.

Ex: Weight, height, number of sales

You can say someone made 2x more sales than someone else and also “30 more sales” than someone else, etc

What measurement of central tendency is better for normal distribution?

Mean

What measurement of central tendency is better for skewed distribution?

Median

What is variance and what does the height of the curve say?

Higher curve= low variance (peanut butter example)

Lower curve= higher variance

Measures the distribution and how much the data varies from the mean.

SD= each value subtracted by the mean and then square the whole thing

What is the significance of standard deviation?

Most people will fall into 1 sd of the mean. One side is 34%, 95% will fall within 2 SD’s

think of the chopped shit on the graph when u chop that shyt up

What is correlation coefficient?

Shows how strong a correlation is on the graph.

0.1-0.3 is most common in psych

What does the equation for linear regression?

Y= B0 + Bx1

OR

y=b+mx

B0 OR mx is the intercept, mx is slope

What is the equation for multiple linear regression

Y= B0 + b1×1 + b2×2 + e

Doing 2 variables to control for confounding factors

ex: b1×1 is job satisfaction, b2×2 is org commitmen

What is a residual in regression analysis?

Observed-predicted

sum of square errors : sigma(y-y²)²

What does a low SSE for a regression analysis mean?

Indicates that the predicted values are close to the actual values, suggesting a good fit

What does a high SSE for a regression analysis mean?

Indicates a poor fit, as the predictions from the actual data points

What is the coefficient of Determination? (R²)

Variance of y/outcome variable accounted for by x/independent variable

What is statistical significance and how is it typically determined?

It is used to determine whether findings are due to coincidences or actually due to real effects. It is typically determined by calculating the p-value

What is p-value?

P-val = P ( |r| > r — H_0 = true)

It helps determine statistical significance

P<0.05 means you can reject the null hypothesis and it IS statistically significant

P>0.05 means you cannot reject the null hypothesis and it IS NOT statistically significant

Can something be valid but not reliable?

No, it has to be reliable. However, something can be reliable but not valid.

What are the 3 main types of validity and their definition?

True correlation:

Discriminative/Divergent: 2 measures of different construct r<0.7

Convergent: 2 measures of same construct, r>0.7

What are the 2 types of ways to study criterion-related validity?

Concurrent: get both x and y at same time

Predictive: measure x and then y later to predict outcome for the future.

ex: measure x now, measure y 6 months later

What is range restriction and how does it affect correlation?

When you only focus on a specific part of the regression model. Makes correlation weaker because you have less data.

What is a meta-analysis? What problems can meta-analysis correct for?

This term is a study that combines all the results from previous studies. Some problems that this can correct include sample size, publication bias, validity, and conflicting results

How is convergent validity different from divergent validity?

Convergent is the same construct with strong validity while Divergent is different constructs with a weak validity

Basically:

Convergent is strong (r>0.7), same constructs

Divergent is weak (r<0.7), different constructs

What is factor analysis?

Statistical technique that organizes large amount of data into understandable and more coherent groups

Ex: examining a big group of data that has the words “Happy, cheerful, and excited" I would then shrink this part of the data into a group called “positive thoughts".”

What is meant by factor loadings? How are factor loadings similar to the correlation coefficient?

It is the relationship between the factor/construct and indicators. It is similar in a way it is similar in size and meaning and ranges from [-1,1]

Ideally between [0.4,0.8]

What are the 3 types of information needed to use a Taylor-Russell table?

Base rate, criterion-validity coefficient, and selection ratio

What does base rate mean?

Percentage of a given population that is successful at a job under a current or previous selection system.

What is validity coefficient? (rxy)

The validity between the predictor/measure and outcome

What is selection ratio and how do you calculate it?

A metric that measures the selectivity of the hiring or admissions process

To calculate it, divide the number of hires by the total number of applicants; for example, if 10 people are hired out of 200 applicants, the selection ratio is 5% or 0.05

What is consequence validity?

Concerned with real world consequences of testing

for example: cognitive ability tests. Intended consequence selects most qualified applicants but unintended consequence discriminate against minorities if the test if culturally-biased

What are 3 main ways to evaluate recruitment effectiveness?

Cost per hire, time to hire, and quality of hire

How to calculate yield ratio?

Hella ratios.

Ex: 100 applicants, 20 get interview, yield ratio of 20%

Ex2: Out of the 20 interviews, 5 people get in = yield ratio of 25%

What are the strengths and weaknesses of internal recruitment?

Strengths: Faster start-up time, lower cost,

Weaknesses: May limit innovations

What is the benefit of external recruitment? 

New ideas, maintaining applicant interest

What does the research say about internal vs. external recruitment?  

Bidwell (2011) found that external recruits have higher levels of experience and education and are paid 18% higher starting salaries

Internal recruits have significantly better retention and better performance

How do researchers conceptualize Job Satisfaction (a multi-faceted construct)? 

Pay, promotions, supervisions, coworkers, and the work itself

What is affective commitment?

People have positive feelings about this org.

“I want to stay”

What is continuance commitment?

People associate high costs with leaving the org

“I need to stay”

What is normative commitment?

People feel obligated to stay with the org

“I oughta stay”

What is job embededness? How does it differ from job satisfaction and organizational commitment?

JE reflects a broad constellation of forces that determine employee retention including LINKS (networks), FIT (how your values fit with the org), SACRIFICE (what you give up leaving the org)

Focuses on the reasons that determine employee retention rather than how much employees want to stay/how much employees like their job

What is a Realistic Job Preview (RJP) and its benefits?

A recruitment tool used to inform applicants of the good and undesirable aspects of a job

Benefits include positive outcomes such as lower expectations, increased p erformance, and higher job satisfaction

How can orgs best conduct diversity recruiting? What are some “dos” and “don’ts”?

Recomment diversity candidates from their networks and encourage applications from minority candidates but do not express preference for a specific minority

How does the ASA theory explain organizational homogeneity over time?

Organization attracts people similar to its values and work culture. Recruits that have different org culture usually leave

How is social media used in hiring and is it recommended to use social media for recruitment?

Orgs practice reviewing social media info when recruiting and assessing job applicants. However, it should be used at a minimum as there is low correlation with job performance and social media content

What is a predictor in personnel psychology?

Measure whose scores represent individual differences on KSAOs linked to effective criterionperformance, whihc is used as the basic for selection or promotion

What is psychological fidelity?

The extent to which an assessment requires test-takers to use KSAOs during the assessment that they will use on the job

What is physical fidelity?

The extent to which an assessment replicates actual tasks performed on the job

What are proximal predictors and distal predictors? How are they different?

The first term assesses KSAOs most directly related to performance (state-like, changes)

The second term is related to criterion performance/job performance but not as immediately as the first term (trait like, stable)

What makes a good predictor?

High criterion-related validity, low subgroup differences

What are compile predictors?

Maximally predictive of job performance. They predict performance better than homogenous measures.

Ex: Service orientation= agreeableness + emotional stability + conscientiousness

What is a compensatory model in selection systems?

Candidates are required to achieve a specfiic total score across ALL assessments but a good score in a certain test can compensate for a lower score on another test

What are the 4 different types of non-compensatory models in selection? And how do they dwerk?

Top-down/rank order: give test, select candidates with top scores based on openings

Cut-off score: applicants must score above cut-off score to BE CONSIDERED

Multiple cutoffs: multiple predicots that each have a cut-off score. Applicants must take all tests and meet/exceed the cut-off score to BE CONSIDERED

Multiple hurdles: applicants must meet minimum cutoff for each predictor BEFORE going onto the next. Failure at any stage means getting dropped from consideration