rescaleTrajectories: Function for rescaling achievement

In mcbeem/ZPDGrowthTrajectories: Simulate achievement growth trajectories using a theoretical model of academic achievement based on the zone of proximal development

Description

rescaleTrajectories performs linear rescaling of the synthetic achievement values produced by ZPDGrowthTrajectories

Usage

rescaleTrajectories(
  trajectories,
  intercept = NULL,
  slope = NULL,
  benchmarks = NULL,
  version = NULL
)

Arguments

trajectories	An object of class ZPD produced by the ZPDGrowthTrajectories() function. If needed, this object will be converted internally to "long" format.
intercept	Optional scalar value providing the intercept for rescaling. Either intercept and slope or benchmarks must be specified. Defaults to NULL.
slope	Optional scalar value providing the slope for rescaling. Either intercept and slope or benchmarks must be specified.Defaults to NULL.
benchmarks	Optional numeric matrix with two columns. The first column contains time points. The second column contains target values for the mean achievement at those time points. The matrix must have at least two rows. Either intercept and slope or benchmarks must be specified. Defaults to NULL.
version	Optional numeric vector indicating that the mean achievement at the target time points should be computed only for students receiving specific version(s) of the curriculum. For example, c(1,2) indicates that the rescaling should be based only on those students receiving versions 1 or 2 of the curriculum. If provided, benchmarks must also be specified. Defaults to NULL.

Details

rescaleTrajectories works in two modes. In the first, mode, the user supplies the intercept and slope values which are used to linearly rescale the synthetic achievement values computed by ZPDGrowthTrajectories. In the second mode, the user supplies a matrix of time points and desired values for the mean achievement at those time points. The function will calculate and apply the optimal rescaling coefficients and report the r-squared value. The user may optionally designate that this transformation should be based on a subset of the versions of the curriculum.

Value

An object of class "ZPD", "data.frame"

See Also

ZPDGrowthTrajectories for simulating growth trajectories.

Examples

# learning rate
learn.rate <- c(.08, .10, .12, .18)

# decay rate
decay.rate <- c(.04, .03, .02, .01)

# initial achievement
initial.ach <- rep(0, times=4)

# quality of home environment
home.env <- c(.06, .12, .15, .20)

# assignment object simulating starting kindergarten on time 801
#  Kindergarten for 200 days, followed by 100 days of summer
#  then 200 days of first grade
assignment <- c(rep(0, times=800), rep(1, times=200),
                rep(0, times=100), rep(2, times=200))

# define school curriculum
curriculum.start.points <- list(
    # "typical curriculum" start points for K and first grade
  matrix(c(.2, .26), nrow=2, ncol=1),
    # "advanced curriculum" start points for K and first grade
  matrix(c(.22, .29), nrow=2, ncol=1)
)

curriculum.widths <- list(
  # "typical curriculum" widths for K and first grade
  matrix(c(.04, .04), nrow=2, ncol=1),
  # "advanced curriculum" widths for K and first grade
  matrix(c(.05, .05), nrow=2, ncol=1)
)

curriculum.review.slopes <- list(
  # "typical curriculum" review slopes for K and first grade
  matrix(c(30, 30), nrow=2, ncol=1),
  # "advanced curriculum" review slopes for K and first grade
  matrix(c(60, 60), nrow=2, ncol=1)
)

curriculum.advanced.slopes <- list(
  # "typical curriculum" advanced slopes for K and first grade
  matrix(c(50, 50), nrow=2, ncol=1),
  # "advanced curriculum" advanced slopes for K and first grade
  matrix(c(25, 25), nrow=2, ncol=1)
)

# students 1 and 2 get typical curriculum, 3 and 4 get advanced
which.curriculum <- c(1,1,2,2)

y <- ZPDGrowthTrajectories(learn.rate=learn.rate, home.env=home.env,
                           decay.rate=decay.rate, initial.ach=initial.ach,
                           ZPD.width=.05, ZPD.offset=.02,
                           home.learning.decay.rate=6,
                           curriculum.start.points=curriculum.start.points,
                           curriculum.widths=curriculum.widths,
                           curriculum.review.slopes=curriculum.review.slopes,
                           curriculum.advanced.slopes=curriculum.advanced.slopes,
                           assignment=assignment, dosage=.8,
                           adaptive.curriculum=FALSE,
                           which.curriculum=which.curriculum,
                           school.weight=.5, home.weight=1, decay.weight=.05,
                           verbose=TRUE)

describeTrajectories(y, byVersion=FALSE)

y2 <- rescaleTrajectories(y, intercept=100, slope=50)
describeTrajectories(y2, byVersion=FALSE)

benchmarks <- matrix(c(
  300, 150.0,
  600, 187.2,
  800, 195.2), ncol=2, byrow=TRUE)

y3 <- rescaleTrajectories(y, benchmarks=benchmarks)
describeTrajectories(y3, byVersion=FALSE)

mcbeem/ZPDGrowthTrajectories documentation built on May 18, 2020, 2:04 p.m.