Lecture 3 Flashcards (1)

Flashcards generated from lecture notes to help students review key concepts for an upcoming exam.

Why is it important to clean data sets before analysis?

Data must be cleaned to ensure accurate results and prevent project failure.

What is the best practice regarding raw data sets and data manipulation?

Retain one untouched data set for reference, and use a copy for manipulation and cleaning.

Why should you remove data from participants who indicated they do not want to participate?

Removing data from participants who did not consent ensures the data reflects only willing participants.

What are two types of respondents whose data may not be reliable?

Straight liners or invariant respondents and rushers.

Why are straight liners considered unreliable respondents?

They provide unreliable data due to a lack of attention, which affects the accuracy of results.

How can you identify 'rushers' in a data set?

Create a duration variable by calculating how long respondents took to complete the survey.

What is the ideal time frame for survey completion to ensure quality responses?

Design surveys to take between 15 and 25 minutes to maintain respondent engagement and data quality.

For a 15-25 minute survey, how long should a participant spend to be considered attentive?

At least five minutes.

In missing data, what are the two types of missing data described?

Participants skip questions they see (indicated by a specific code) and those who don't see the question (appearing as blank).

Why must missing values be defined in SPSS?

To ensure they are not included in calculations, which would skew the average income or other metrics.

How do you create a respondent ID in SPSS?

Create a new variable using the $CASENUM system command.

What is the 'trick' to placing a new variable in a specific column in SPSS?

Insert a variable where you want the new variable to appear.

Why are variables sometimes reverse coded in surveys?

To ensure participants are paying attention and providing thoughtful answers.

When reverse coding, is it better to recode into the existing variable or into a new variable?

Recode the item into a new variable to ensure the original data is preserved in case of errors.

What does mean centering achieve?

Balancing the variable around a neutral value by subtracting the mean.

Why is standardizing variables important?

It normalizes variables, making them directly comparable despite different scales, using the formula (x - mean) / standard deviation.

What is the purpose of dichotomizing continuous data?

Client wants data only for people over or under a certain age.

What action can you take if you want to analyze an age range ordinally?

Recode to impute a numeric value for analysis where the range is creating an issue.