Quantitative Analysis/ Structural Equation Modelling
Measurement and effects together

Structural Equation Modelling: How to test a whole chain of ideas, from survey topics to outcomes, in one model

A structural equation model measures topics such as quality or trust from several survey questions and tests how the topics affect each other, all in one model. This guide covers the ideas, the counting, the calculations, every panel of the output, a worked example and how to report it.

24 min read Structural equation modelling Advanced

01What is a structural equation model?

A structural equation model tests how topics that you measure with several questions affect each other. It joins two ideas. The first is measuring a hidden topic, such as trust, from a set of questions. The second is asking which topic influences which.

It is often shortened to SEM. The name sounds heavy, but the question it answers is plain: if one topic goes up, what happens to the others, and through which route?

An example

You surveyed 500 people in a host community. Eleven questions cover three topics: service quality (four questions), trust in providers (four questions) and use of services (three questions, such as how likely people are to go back). You think good service builds trust, and trust makes people use the services. You want to know three things. Does quality raise use? Does it do so only by building trust, or also directly? And how much of the differences in trust and use do these topics explain?

Why not just average the questions and run a regression?

You can, and it is a fine first look. But an average of questions still carries the noise of each question. That noise weakens the links between topics, so the effects look smaller than they are. A structural equation model separates the topic from the noise before it estimates the links. It also lets one topic be the result of one relationship and the cause in the next, all in one model.

What you get out of it

  1. Fit
    Whether the model as a whole matches your data.
  2. Measurement check
    Loadings, reliability and whether topics are separate, as in a confirmatory factor analysis.
  3. Paths
    The size and direction of each link between topics, with intervals.
  4. Routes
    The direct, indirect and total effect of one topic on another, route by route.
  5. Variance explained
    How much of each outcome topic the model accounts for.

02When to use it

Use it when your key variables are scores built from several questionnaire items and you have a clear idea of how they connect. Typical cases are attitudes, trust, satisfaction, well-being and similar topics.

If all your variables are single measured numbers, such as age, income or a test score, use path analysis. If you only want to know whether one variable carries the effect of another, mediation analysis is simpler. If you only want to check that questions group as planned, use confirmatory factor analysis.

You need a model idea before you start, and a sample that is large enough. The module asks for at least 200 cases, and more for bigger models.

03What you need before you start

Requirements and what happens if they are not met
PointWhat is neededIf it is not met
Sample sizeAt least 200 cases, and more when the model has many topics, few questions per topic or weak loadings. As a floor, 10 cases for every number the model estimates.Estimates and fit become unstable. Standard errors are too wide to be useful.
Topics and questionsEach topic has at least two questions, and three or more is safer. Each question belongs to one topic.The module stops with a message.
A model ideaThe arrows come from theory or from the design of the programme, not from what the data happen to show.The result describes a model built to fit, and will probably not hold on new data.
Direction of arrowsEach arrow says one thing leads to another. The data cannot confirm the direction. Only your design and logic can.A model with the arrows reversed often fits equally well.
No loopsThe module fits models where effects run one way. Two topics cannot each affect the other.The module stops, or the model cannot be estimated.
Reverse-worded questionsRecode first, so a high score always means more of the topic.Negative loadings and poor reliability.
Missing answersA case missing any variable in the model is left out. The cases used are shown.Check missing values first if many cases drop out.

04Two parts in one model

A structural equation model has a measurement part and a structural part.

  1. The measurement part says which questions belong to which topic. It is the same as a confirmatory factor analysis.
  2. The structural part says which topic affects which. These are the arrows between the ovals.

Figure 1 shows the structural part of the example. Solid arrows are links you expect. The dashed arrow is a direct link from quality to use that skips trust. You include it to find out whether trust really carries the whole effect.

Figure 1. The model in the exampleThree ovals in a row: service quality, trust in providers and use of services. A solid arrow runs from quality to trust and from trust to use. A dashed arrow runs from quality directly to use.Service quality(4 questions)Trust in providers(4 questions)Use of services(3 questions)0.600.560.07
Figure 1. The model in the example, with standardised path values

The numbers on the arrows are standardised, so they run roughly from -1 to 1 and can be compared. A value of 0.60 means that people who are one standard deviation higher on quality are about 0.6 of a standard deviation higher on trust.

05How it is calculated

One model, estimated together

The software estimates the measurement part and the structural part at once. As in confirmatory factor analysis, it picks the loadings, error sizes and path values that make the correlations predicted by the model as close as possible to the correlations in your data. Estimating together is what lets it separate noise from the topics.

Direct, indirect and total effects

An effect can travel in more than one way. Quality can reach use directly, and it can reach use through trust. The indirect effect is the product of the arrows along the route. The total effect adds the direct and indirect parts.

indirect effect = path(quality to trust) × path(trust to use)
total effect = direct effect + indirect effect

With the unstandardised values from the example: 0.686 × 0.436 = 0.299 for the indirect effect. The direct effect is 0.063, so the total is 0.063 + 0.299 = 0.362. In standardised terms the indirect effect is 0.599 × 0.558 = 0.33.

Variance explained

Trust is predicted only by quality, so the variance explained is the square of the standardised path: 0.599² = 0.36. About 36% of the differences in trust are accounted for by quality. For use, quality and trust together account for about 36%.

Comparing two versions of a model

To test whether a path is needed, fit the model with and without it and compare the chi-square values. Without the direct path, the model fits with a chi-square of 24.3 on 41 degrees of freedom. With it, the chi-square is 23.3 on 40. The difference is 1.0 for one extra path, which is small. Adding the direct path barely improves the model.

06The model in more detail

The two sets of equations

A structural equation model is two models joined together. The measurement part says each question is a noisy reading of a topic. The structural part says how the topics affect each other.

measurement: answer to a question = loading × topic + error
structural: outcome topic = path × predictor topic + path × another predictor + residual

The topics are never observed. The model finds the topic values that best explain the answers, and it estimates the loadings and the paths at the same time. This is why the paths are not weakened by the noise in the questions. If you average the questions and run a regression, the noise in the averages pulls the paths toward zero.

Counting what is estimated

The example has 11 questions: 4 on quality, 4 on trust and 3 on use. That gives 11×12/2 = 66 variances and covariances for the model to explain. The model estimates 26 numbers: 8 loadings (the first question of each topic is fixed to set the scale), 11 error variances, 3 paths, 1 variance for the quality topic, 2 residual variances for trust and use, and 1 correlation between the errors of two quality questions (the change suggested in the confirmatory analysis). The difference, 66 - 26 = 40, is the degrees of freedom. It counts the independent ways the model can disagree with the data. The paths you leave out, and the questions you do not let load on other topics, are what make the model testable. A model in which every topic is linked to every other, and every question to every topic, would fit perfectly and tell you nothing.

If a model estimates exactly as many numbers as there are variances and covariances, the degrees of freedom are zero and the fit is perfect by design. If it estimates more, it cannot be estimated, and the module tells you to simplify it.

How the numbers are found

Estimation works as in confirmatory factor analysis. For numeric questions the module uses maximum likelihood. It searches for the loadings and paths that make the variances and covariances implied by the model as close as possible to the real ones, and the chi-square is the leftover gap multiplied by the number of cases. For ordered or yes or no questions it uses polychoric correlations and weighted least squares with an adjusted chi-square. With a survey weight, estimates use the weights and standard errors follow the survey design.

Effects through the model

When a topic affects another through a middle topic, the indirect effect is the product of the arrows along the route. The total effect is the direct effect plus all indirect effects.

indirect effect of Quality on Use = path(Quality to Trust) × path(Trust to Use)
total effect = direct effect + indirect effect

In the example the standardised paths are 0.60 and 0.56, so a rough check gives 0.60 × 0.56 = 0.34. The table shows the exact indirect effect, which is worked out from the unrounded estimates. The direct path from Quality to Use is 0.07 and its interval includes zero, so the total effect of 0.362 is almost all indirect. Intervals for these products are worked out with the delta method for numeric variables.

Fit

The fit measures are those used in confirmatory factor analysis: chi-square, CFI, TLI, RMSEA with its range, and SRMR. Here the model has chi-square 23.3 on 40 degrees of freedom, CFI 1.000, RMSEA 0.000 and SRMR 0.020. Good fit has CFI and TLI of 0.95 or more, RMSEA of 0.06 or less, and SRMR of 0.08 or less. Read all of them together.

Variance explained

For each outcome topic, the module gives the share of its variation that the topics leading to it account for, as a percentage. It tells you how much your model explains, not whether it is right.

Comparing two versions of a model

To test whether an arrow is needed, run the model with and without it. The model with fewer arrows has more degrees of freedom. If its chi-square is not much worse, the simpler model is enough. The AIC and BIC columns help compare models that are not nested. Lower values are better, and BIC penalises extra arrows more strongly.

07Reading the output

Model fit

Read the fit as in a confirmatory factor analysis. CFI and TLI of 0.95 or higher, RMSEA of 0.06 or lower and SRMR of 0.08 or lower are good. A model that fits badly cannot be trusted, even when its paths look strong. Check fit first.

Loadings, reliability and separation

The measurement part is reported as in confirmatory factor analysis: standardised loadings, composite reliability, average variance extracted, and the heterotrait-monotrait ratio for separation. A weak measurement part makes the structural results hard to trust.

Structural paths

The paths table gives each link with its raw value, standardised value, interval and p-value. An interval that includes zero means the link could be zero. Standardised values let you compare links within the same model.

Effects route by route

For each pair of topics with more than one route, the module separates the direct, indirect and total effect, with intervals. It also lists each indirect route on its own, so you can see which one carries the effect.

Variance explained

This shows how much of each outcome the model accounts for. There is no universal good value. It depends on the field.

Suggested changes

As in confirmatory factor analysis, the module suggests changes that would improve fit. Add one only if you can explain it in plain words.

Types of answers and of other predictors

For items that are yes or no, or only a few ordered categories, choose the mix of binary, ordinal and continuous. Predictors that are not factors, such as sex or age group, can be added as ordinary variables, and you say whether each is numeric, binary or categorical.

08How to run it in AnalyZ Solutions

Open Structural Equation Modelling and choose Structural Equation Model. The module is marked Beta.

  1. Define the factors
    Name each factor and choose its questions, as in a confirmatory factor analysis. Each factor needs at least two.
  2. Say how answers are treated
    Choose all continuous, or the mix of binary, ordinal and continuous.
  3. Define the structure
    Add one equation for each outcome. Pick the outcome, then its predictors. Predictors can be factors or ordinary variables.
  4. Click Run Structural Equation Model
    You get the fit, the loadings, reliability, the paths, the effects route by route, the variance explained, suggested changes and a plain-language interpretation.
  5. Review the suggestions
    Add a change only if you can justify it. The model re-runs with it.
When your data come from a survey with sampling weights. With a survey weight set, the model is estimated with the weights and the weighted number of cases is shown next to the sample size.

09Reading the output, panel by panel

Every panel in the Structural Equation Modelling output
PanelWhat it showsWhat to look for
Model fitChi-square, df, CFI, TLI, RMSEA with its range, SRMR, AIC, BIC.Several measures in the good range, not one.
Loadings, reliability and separationAs in confirmatory factor analysis: loadings, composite reliability, AVE, factor correlations, discriminant validity.Fix the measurement part first. Weak loadings weaken everything built on them.
Structural pathsEach arrow between topics or variables, raw and standardised, with intervals and p-values.The direction, size and interval of each arrow. An interval that includes zero means no clear effect.
Effects route by routeDirect, indirect and total effects, route by route, with intervals.Whether the effect goes mostly through the middle topic.
Variance explainedThe share of each outcome explained.Low values mean other things matter more than the model includes.
Suggested changesModification indices for paths and covariances left out.Add a change only if you can explain it in words.
InterpretationA plain-language summary.A first draft. Check it against the tables.
Messages you may see
MessageWhat it meansWhat to do
The model is not identifiedThe data do not contain enough information for all the numbers.Add questions to topics with only two, or remove arrows.
More estimates than the data can supportThe model has more numbers than the data have variances and covariances.Simplify the model.
An error variance is at or near zeroA question is almost the same as its topic.Check the question. Add a third question to the topic.
A standardised loading is above 1An improper solution, often from a small sample or too few questions.Simplify the model. Check for near-duplicate questions.
The model did not fully convergeThe search did not finish cleanly.Treat results with caution. Simplify or add cases.

A sensible order of work

  1. Check each topic on its own
    Run confirmatory factor analysis on the questions. If a topic does not hold together, no structural result can be trusted.
  2. Fit the whole model
    Add the arrows your theory says exist.
  3. Read fit, then loadings, then paths
    Do not interpret arrows from a model that fits poorly.
  4. Read the effects route by route
    Use the indirect and total effects. Do not add up arrows by hand.
  5. Consider a rival model
    If the arrows could run another way, fit that version too, and say what you found.

10Interpreting the result

The figures below are invented to show the steps. They come from a real run of the module on a simulated survey of 500 people, with the questions described earlier.

Model fit
Chi-squaredfp-valueCFITLIRMSEA (90% CI)SRMR
23.3400.9841.0001.0110.000 (0.000 to 0.000)0.020

Step 1: check the fit. CFI is 1.000, RMSEA is 0.000 and SRMR is 0.020, so the model fits the data well. A fit this clean is unusual in real surveys. With real data, expect some compromise.

Measurement part
FactorComposite reliabilityAVEWeakest loading
Service quality0.790.480.66
Trust in providers0.820.540.55
Use of services0.750.500.69

Step 2: check the measurement part. Composite reliability is 0.75 or higher for every factor, and the weakest single loading is 0.55. The topics are measured well enough to interpret the paths.

Structural paths
PathRawStandardised95% CIp-value
Service quality → Trust in providers0.6860.600.55 to 0.82< 0.001
Trust in providers → Use of services0.4360.560.32 to 0.55< 0.001
Service quality → Use of services (direct)0.0630.07-0.06 to 0.190.321

Step 3: read the paths. Quality has a clear effect on trust (standardised 0.60). Trust has a clear effect on use (0.56). The direct path from quality to use is small (0.07) and its interval, -0.06 to 0.19, includes zero.

Effect of service quality on use of services
EffectRaw95% CIStandardised
Direct0.063-0.06 to 0.190.07
Indirect, through trust0.2990.20 to 0.390.33
Total0.3620.25 to 0.470.41

Step 4: read the effects. The total effect of quality on use is 0.36. Almost all of it, 0.30, goes through trust. The direct part is small and cannot be told from zero. Roughly 83% of the total effect travels through trust.

Step 5: look at variance explained. Quality accounts for about 36% of the differences in trust. Quality and trust together account for about 36% of the differences in use.

In plain words

In this survey, people who rated services as better trusted providers more, and people who trusted providers more said they would use the services more. Trust appears to carry nearly all of the link between service quality and use. Once trust is taken into account, quality adds almost nothing on its own. The model describes patterns in this sample. It does not prove that improving quality would raise trust.

11How to report the result

Example write-up. A structural equation model was fitted to the answers of 500 respondents using maximum likelihood. The model linked service quality to trust in providers and trust to use of services, with a direct path from quality to use. Fit was good (CFI = 1.000; RMSEA = 0.000; SRMR = 0.020). Quality was clearly linked to trust (standardised path 0.60) and trust to use (0.56). The direct path from quality to use was small (0.07) and its interval included zero. The total effect of quality on use (0.36) was therefore almost entirely indirect through trust (0.30). These results describe associations that are consistent with the proposed model. They do not show cause, because the data were collected at one time.

12A checklist before you trust the result

  1. The model comes from theory or design, and you can say why each arrow is there.
  2. Each topic was checked on its own first.
  3. You have at least 200 cases and at least 10 per estimated number.
  4. Fit is good on several measures.
  5. Loadings, reliability and discriminant validity are acceptable.
  6. You read indirect and total effects from the effects panel.
  7. Every change you added has a reason and is reported.
  8. You have thought about a rival model with arrows the other way.
  9. Your wording says "linked to" or "consistent with", not "caused", unless the design supports it.

13Common mistakes

Reading arrows as proof of cause. The model shows patterns that agree or disagree with your idea. Data collected at one point in time cannot show which way influence runs.

Skipping the measurement check. If a topic is poorly measured, every path that involves it is unreliable.

Building a model with no theory. Trying many arrow patterns until one fits produces a model that rarely holds up.

Too complex a model for the sample. More factors and paths need more cases. If the model will not converge or gives odd values, simplify it.

Ignoring other models that fit equally well. A model that fits does not rule out others that fit as well. Say so in your report.

Judging a path by its p-value alone. Look at its size and interval as well.

Frequently Asked Questions
What is the difference between structural equation modelling and path analysis?
Path analysis uses variables you measured directly. A structural equation model also includes topics measured from several questions and corrects for the noise in those questions.
How many cases do I need?
A common guide is at least 200, and more for larger models. Models with many factors, few questions per factor or weak loadings need more.
Can I include a variable like sex or age group?
Yes. Add it as a predictor in an equation and say whether it is numeric, binary or categorical.
The direct path is not significant. Should I remove it?
Not automatically. Keep it if your theory says it could exist. Compare the model with and without it, and report what you find.
Does a good fit mean my model is right?
No. It means the model is consistent with the data. Other models might fit as well.
Can I use yes or no questions or short rating scales?
Yes. Choose the mix of binary, ordinal and continuous so the analysis treats them as ordered categories.
Why not just average the questions and run a regression?
Averages carry the noise of the questions, and that noise pulls the estimated effects toward zero. A structural equation model separates the topic from the noise, so the paths are not weakened by measurement error.
Can the data tell me which way an arrow points?
No. A model with an arrow reversed can fit as well as the original. The direction comes from your theory, from timing in the data, or from the design of the study.
What are degrees of freedom here?
The number of variances and covariances in the data minus the number of values the model estimates. It is the number of independent ways the model can fail to match the data.
Why do I need at least 200 cases?
The model estimates many numbers at once. With fewer cases, the estimates are noisy and the fit values unreliable.
Should I remove a path whose interval includes zero?
Keep it if theory says it may exist, and report that it was small. Removing it changes the other estimates a little. You can compare both versions.

Test how your topics connect

Open Structural Equation Modelling in AnalyZ Solutions. Free to sign up, free to analyze.

Get started
Related Guides