What is structural equation modelling?

	a)	A set of statistical procedures that may be applied to quantitative data.
	b)	Tests models of the relationship between observed and unobserved variables.
	c)	Tests the relationship between sets of latent variables.
	d)	All of the above.
Correct! For more information about this topic see section 21.1. Sorry that answer is incorrect, please try again. See section 21.1 for more information.Your answer has been saved.

In which way does structural equation modelling differ from other statistical procedures?

	a)	You have to make explicit your theories and models of the relationships between variables.
	b)	'Fishing' or data-mining is relatively easy to do with SEM.
	c)	The process is less rigorous than most analytical procedures.
	d)	There are a very limited number of software packages.
Correct! For more information about this topic see section 21.2. Sorry that answer is incorrect, please try again. See section 21.2 for more information.Your answer has been saved.

Which of the following statements is true?

	a)	In SEM, if the model you have proposed is not consistent with the data, it must be re-analysed.
	b)	In SEM, if the model is not consistent with the data, it must be rejected.
	c)	In SEM, if the model is consistent with the data it means that you have found the true model.
	d)	In SEM, the idea is to generate theoretical models from data.
Correct! For more information about this topic see section 21.2. Sorry that answer is incorrect, please try again. See section 21.2 for more information.Your answer has been saved.

When analysing data, each time a new parameter is added...

	a)	Accuracy of the model will increase or remain constant.
	b)	It is not possible to add new parameters to the model.
	c)	The predictive power of the test is reduced.
	d)	It complicates the procedure and makes the data more difficult to interpret.
Correct! For more information about this topic see p.448. Sorry that answer is incorrect, please try again. See p.448 for more information. Your answer has been saved.

Define the term 'error'

	a)	The residual variance.
	b)	Errors made by participants during the testing session.
	c)	Errors made by the researcher during the testing session.
	d)	Errors made by the researcher whilst analysing data.
Correct! For more information about this topic see p.449. Sorry that answer is incorrect, please try again. See p.449 for more information.Your answer has been saved.

When conducting SEM, there are a number of decisions that have to be made. Consider the following and decide which decision is the most important.

	a)	Whether the model is consistent with the data.
	b)	Whether the model makes good enough predictions.
	c)	Whether making the model more complex has significantly improved its predictive power.
	d)	All of the above.
Correct! For more information about this topic see p.450. Sorry that answer is incorrect, please try again. See p.450 for more information.Your answer has been saved.

When developing a model, the researcher must consider the following...

	a)	How many latent variables there are and how they correlate with each other.
	b)	Which observed variables are affected by which common factor or latent variable?
	c)	Which observed variables are affected a unique factor and do these pairs of factors correlate?
	d)	All of the above.
Correct! For more information about this topic see p.451. Sorry that answer is incorrect, please try again. See p.451 for more information.Your answer has been saved.