FAME Bundle 4 – Courses 7 and 8

ACER's Foundation in Applied Measurement in Education (FAME) is a suite of 8 short, online, modular courses.

Rasch modelling short courses cover:

• introductory concepts and theoretical understanding of educational assessment and measurement
• item and test analysis
• differential item functioning
• test equating
• psychometric analysis reports
• large-scale analysis of educational data

COURSES 1 AND 2 Focus on the theoretical foundations that are essential to large-scale educational assessment.

COURSES 3 TO 8 Focus on applying the skills and require ACER ConQuest® software. Course 7 and 8 is offered for advanced analysis of large-scale education data. For more information about ACER ConQuest®, please go to the ACER ConQuest webpage.

FAME Bundle 4 includes Courses 7 and 8:

FAME 7 Working with plausible values (Advanced / FAME 1–6 / Experienced analysts.  2 days. ACER ConQuest®, conquestr and RStudio software)

• Introduces the theory behind different item response models and case ability estimation routines that can be used to estimate population parameters reliably.
• Participants will apply this knowledge to fit a combined item response and population model, similar to those seen in large-scale assessment programs such as NAPLAN and the Programme for International Student Assessment, and to correctly apply the law of total variance to undertake secondary analysis to yield unbiased population parameter estimates.

FAME 8 Complex survey designs in large-scale assessment (Advanced / FAME 1–6 / Experienced analysts. 2 days. ACER ConQuest®, conquestr and RStudio software)

• Introduces complex survey designs used in large-scale assessment programs to accurately estimate population parameters that are representative of national or sub-national populations of interest.
• Participants will apply this knowledge to correctly apply the law of total variance to undertake secondary analysis to yield unbiased population parameter estimates using plausible values, and in addition using replicate weights to perform re-sampling to account for sampling error.