Alber-Morgan, et al. (2007)

Description

Data from a multiple baseline design conducted by Alber-Morgan, Ramp, Anderson, & Martin (2007). The variables are as follows:

• case Participant identifier

• condition Factor identifying the phase of the design (baseline or treatment)

• session Measurement occasion

• outcome Number of words read correctly per minute

Format

A data frame with 119 rows and 4 variables

Source

Alber-Morgan, S. R., Ramp, E. M., Anderson, L. L., & Martin, C. M. (2007). Effects of repeated readings, error correction, and performance feedback on the fluency and comprehension of middle school students with behavior problems. Journal of Special Education, 41(1), 17-30. doi:10.1177/00224669070410010201

Example 2 from Hedges, Pustejovsky, & Shadish (2012)

Description

Data from an ABAB design conducted by Anglesea, Hoch, & Taylor (2008). The variables are as follows:

• case Case identifier.

• condition Factor indicating baseline or treatment condition

• phase Study phase (including both control and treatment condition)

• session Measurement occasion

• outcome Total seconds of eating time

Format

A data frame with 55 rows and 5 variables

Source

Anglesea, M. M., Hoch, H., & Taylor, B. A. (2008). Reducing rapid eating in teenagers with autism: Use of a pager prompt. Journal of Applied Behavior Analysis, 41(1), 107-111. doi:10.1901/jaba.2008.41-107

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2012). A standardized mean difference effect size for single case designs. Research Synthesis Methods, 3, 224-239. doi:10.1002/jrsm.1052

Barton-Arwood, Wehby, & Falk (2005)

Description

Data from a multiple baseline design conducted by Barton-Arwood, Wehby, and Falk (2005). The variables are as follows:

• case Participant identifier

• condition Factor identifying the phase of the design (A or B)

• session Measurement occasion

• outcome Oral reading fluency score (words per minute)

Format

A data frame with 143 rows and 4 variables

Source

Barton-Arwood, S. M., Wehby, J. H., & Falk, K. B. (2005). Reading instruction for elementary-age students with emotional and behavioral disorders: Academic and behavioral outcomes. Exceptional Children, 72(1), 7-27. doi:10.1177/001440290507200101

Bryant et al. (2018)

Description

Data from a clustered multiple baseline across participants design conducted by Bryant et al. (2018). The variables are as follows:

• Study_ID. Study identifier.

• school. School identifier.

• group. Group identifier.

• case. Student identifier.

• treatment. Indicator for treatment phase.

• session. Measurement occasion.

• session_trt. Measurement occasion times treatment phase.

• outcome. Texas Early Mathematics Inventory (TEMI-Aim Check) scores.

• session_c. Measurement occasion centered at the follow-up time.

Format

A data frame with 536 rows and 8 variables

Source

Bryant, D. R., Bryant, B. R., Sorelle-Miner, D. A., Falcomata, T. S. & Nozari, M. (2018). Tier 3 intensified intervention for second grade students with severe mathematics difficulties. Archives of Psychology, 2(11), 1-24. doi:10.31296/aop.v2i11.86

Calculates a confidence interval for a standardized mean difference effect size

Description

Calculates a confidence interval given a g_REML, a g_HPS, or a g_mlm object using either a central t distribution (for a symmetric interval) or a non-central t distribution (for an asymmetric interval).

Arguments

Value

A vector of upper and lower confidence bounds.

Examples

data(Laski)
Laski_RML <- lme(fixed = outcome ~ treatment,
                 random = ~ 1 | case,
                 correlation = corAR1(0, ~ time | case),
                 data = Laski)
Laski_g_REML <- suppressWarnings(
  g_REML(Laski_RML, p_const = c(0,1), 
         r_const = c(1,0,1), returnModel = FALSE)
)
CI_g(Laski_g_REML, symmetric = TRUE)
CI_g(Laski_g_REML, symmetric = FALSE)

Laski_HPS <- with(Laski, effect_size_MB(outcome, treatment, case, time))
CI_g(Laski_HPS, symmetric = FALSE)

Laski_g_mlm <- g_mlm(mod = Laski_RML, p_const = c(0,1), r_const = c(1,0,1))
CI_g(Laski_g_mlm, symmetric = FALSE)

Carson (2008)

Description

Data from a BAB design conducted by Carson, Gast, & Ayres (2008). The variables are as follows:

• case Participant identifier

• treatment Factor describing the treatment condition

• phase Numeric describing the phase of the study design for each case

• outcome Outcome scores

• time Measurement occasion

Format

A data frame with 47 rows and 5 variables

Source

Carson, K. D., Gast, D. L., & Ayres, K. M. (2008). Effects of a photo activity schedule book on independent task changes by students with intellectual disabilities in community and school job sites. European Journal of Special Needs Education, 23, 269-279.

Case, Harris, and Graham (1992)

Description

Data from a multiple baseline design conducted by Case, Harris, and Graham (1992). The variables are as follows:

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. Number of subtraction equations and answers correct on each word problem probe.

Format

A data frame with 56 rows and 4 variables

Source

Case, L. P., Harris, K. R., & Graham, S. (1992). Improving the mathematical problem-solving skills of students with learning disabilities: Self-regulated strategy development. The Journal of Special Education, 26(1), 1-19. doi:10.1177/002246699202600101

Datchuk (2016)

Description

Data from a multiple baseline design conducted by Datchuk (2016). The variables are as follows:

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. Correct word sequences per minute.

Format

A data frame with 74 rows and 4 variables

Source

Datchuk, S. M. (2016). Writing simple sentences and descriptive paragraphs: Effects of an intervention on adolescents with writing difficulties. Journal of Behavioral Education, 25(2), 166-188. doi:10.1007/s10864-015-9236-x

Delemere & Dounavi (2018)

Description

Data from a multiple baseline design conducted by Delemere & Dounavi (2018). The variables are as follows:

• intervention. Type of intervention received: bedtime fading or positive routines.

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. Total sleep onset latency in minutes across nights.

Format

A data frame with 187 rows and 5 variables

Source

Delemere, E., & Dounavi, K. (2018). Parent-implemented bedtime fading and positive routines for children with autism spectrum disorders. Journal of Autism and Developmental Disorders, 48(4), 1002-1019. doi:10.1007/s10803-017-3398-4

Gunning & Espie (2003)

Description

Data from a multiple baseline design conducted by Gunning & Espie (2003). The variables are as follows:

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. Sleep onset latency in minutes.

Format

A data frame with 301 rows and 4 variables

Source

Gunning, M. J., & Espie, C.A. (2003). Psychological treatment of reported sleep disorder in adults with intellectual disability using a multiple baseline design. Journal of Intellectual Disability Research, 47(3), 191-202. doi:10.1046/j.1365-2788.2003.00461.x

Calculate expected information matrix

Description

Calculates the expected information matrix from a fitted linear mixed effects model with AR(1) correlation structure in the level-1 errors.

Usage

Info_Expected_lmeAR1(m_fit)

Arguments

Value

Expected Information matrix corresponding to variance components of m_fit.

Examples

data(Laski)
Laski_RML <- lme(fixed = outcome ~ treatment, 
                 random = ~ 1 | case, 
                 correlation = corAR1(0, ~ time | case), 
                 data = Laski)
Info_Expected_lmeAR1(Laski_RML)

Example 1 from Hedges, Pustejovsky, & Shadish (2012)

Description

Data from an ABAB design conducted by Lambert, Cartledge, Heward, & Lo (2008). The variables are as follows:

• measure. Outcome measure description (disruptive behavior or academic response).

• case. Student identifier.

• treatment. Factor indicating treatment or control condition. SSR = single-subject responding. RC = response cards.

• phase. Study phase (including both control and treatment condition)

• time. Measurement occasion.

• outcome. Outcome scores.

Format

A data frame with 461 rows and 6 variables

Source

Lambert, M. C., Cartledge, G., Heward, W. L., & Lo, Y. (2006). Effects of response cards on disruptive behavior and academic responding during math lessons by fourth-grade urban students. Journal of Positive Behavior Interventions, 8(2), 88-99. doi:10.1177/10983007060080020701

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2012). A standardized mean difference effect size for single case designs. Research Synthesis Methods, 3, 224-239. doi:10.1002/jrsm.1052

Example 2 from Hedges, Pustejovsky, & Shadish (2013)

Description

Data from a multiple baseline design conducted by Laski, Charlop, & Schreibman (1988). The variables are as follows:

• case. Child identifier.

• outcome. Frequency of child vocalization, as measured by a partial interval recording procedure with 60 ten-second intervals per session.

• time. Measurement occasion.

• treatment. Indicator for treatment phase.

Format

A data frame with 128 rows and 4 variables

Source

Laski, K. E., Charlop, M. H., & Schreibman, L. (1988). Training parents to use the natural language paradigm to increase their autistic children's speech. Journal of Applied Behavior Analysis, 21(4), 391-400. doi:10.1901/jaba.1988.21-391

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2013). A standardized mean difference effect size for multiple baseline designs across individuals. Research Synthesis Methods, 4(4), 324-341. doi:10.1002/jrsm.1086

MB1 simulation results

Description

Simulation results for model MB1 from Pustejovsky, Hedges, & Shadish (2014).

Format

A data frame

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

MB1 simulation time

Description

MB1 simulation time

Format

A data frame

MB2 simulation results

Description

Simulation results for model MB2 from Pustejovsky, Hedges, & Shadish (2014).

Format

A data frame

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

MB2 simulation time

Description

MB2 simulation time

Format

A data frame

MB4 simulation results

Description

Simulation results for model MB4 from Pustejovsky, Hedges, & Shadish (2014).

Format

A data frame

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

MB4 simulation time

Description

MB4 simulation time

Format

A data frame

Musser (2001)

Description

Data from a multiple baseline design conducted by Musser, Bray, Kehle, and Jenson (2001). The variables are as follows:

• student Participant identifier

• session Measurement occasion

• outcome Percentage of disruptive intervals

• treatment Factor indicating baseline, treatment, or follow-up phase

Format

A data frame with 136 rows and 4 variables

Source

Musser, E. H., Bray, M. A., Kehle, T. J., & Jenson, W. R. (2001). Reducing disruptive behaviors in students with serious emotional disturbance. School Psychology Review, 30(2), 294-304.

Peltier et al. (2020)

Description

Data from a multiple baseline design conducted by Peltier, Sinclair, Pulos, & Suk (2020). The variables are as follows:

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. Mathematical problem-solving performance (percentage).

Format

A data frame with 232 rows and 4 variables

Source

Peltier, C., Sinclair, T. E., Pulos, J. M., & Suk, A. (2020). Effects of schema-based instruction on immediate, generalized, and combined structured word problems. The Journal of Special Education, 54(2), 101-112. doi:10.1177/0022466919883397

Rodgers et al. (2021)

Description

Data from a multiple baseline design conducted by Rodgers, Datchuk, & Rila (2021). The variables are as follows:

• case. Participant identifier.

• session. Measurement occasion.

• condition. Factor identifying the phase of the design (baseline or treatment).

• outcome. The number of correct writing sequences in 1 minute.

Format

A data frame with 83 rows and 4 variables

Source

Rodgers, D. B., Datchuk, S. M., & Rila, A. L. (2021). Effects of a text-writing fluency intervention for postsecondary students with intellectual and developmental disabilities. Exceptionality, 29(4), 310-325. doi:10.1080/09362835.2020.1850451

Rodriguez & Anderson (2014)

Description

Data from a multiple baseline design conducted by Rodriguez and Anderson (2014). The variables are as follows:

• case Participant identifier

• condition Factor identifying the phase of the design (A or B)

• session Measurement occasion

• outcome Percentage of intervals with problem behavior

Format

A data frame with 148 rows and 4 variables

Source

Rodriguez, B. J., & Anderson, C. M. (2014). Integrating a social behavior intervention during small group academic instruction using a total group criterion intervention. Journal of Positive Behavior Interventions, 16(4), 234-245. doi:10.1177/1098300713492858

Romaniuk (2002)

Description

Data from a treatment reversal design conducted by Romaniuk and colleagues (2002). The variables are as follows:

• case Participant identifier

• phase Factor identifying the phase of the design

• condition Factor identifying the treatment condition

• session Measurement occasion

• outcome Problem behavior

• measurement Character string describing how problem behavior was measured

Format

A data frame with 148 rows and 4 variables

Source

Romaniuk, C., Miltenberger, R., Conyers, C., Jenner, N., Jurgens, M., & Ringenberg, C. (2002). The influence of activity choice on problem behaviors maintained by escape versus attention. Journal of Applied Behavior Analysis, 35(4), 349-62. doi:10.1901/jaba.2002.35-349

Ruiz, et al. (2020)

Description

Data from a multiple baseline design conducted by Ruiz, Luciano, Florez, Suarez-Falcon, & Cardona-Betancourt (2020). The variables are as follows:

• case. Participant identifier.

• measure. Outcome measure description (AAQ-II, ANXIETY, CFQ, DASS-TOTAL, DEPRESSION, PSWQ, PTQ, STRESS, VQ-OBSTRUCTION, or VQ-PROGRESS).

• treatment Factor indicating baseline, treatment, post, or follow-up phase.

• time. Measurement occasion.

• outcome. Outcome scores.

Format

A data frame with 840 rows and 5 variables

Source

Ruiz, F., Luciano, C., Florez, C., Suarez-Falcon, J., & Cardona-Betancourt, V. (2020). A Multiple-Baseline Evaluation of Acceptance and Commitment Therapy Focused on Repetitive Negative Thinking for Comorbid Generalized Anxiety Disorder and Depression. Frontiers in Psychology, 11. doi:10.3389/fpsyg.2020.00356

Example 1 from Hedges, Pustejovsky, & Shadish (2013)

Description

Data from a multiple baseline design conducted by Saddler, Behforooz, & Asaro, (2008). The variables are as follows:

• case Student identifier

• measure Factor indicating the outcome measure (writing quality, T-unit length, number of constructions)

• outcome Value of outcome measure.

• time. Measurement occasion.

• treatment. Factor indicating the treatment phase.

Format

A data frame with 124 rows and 5 variables

Source

Saddler, B., Behforooz, B., & Asaro, K. (2008). The effects of sentence-combining instruction on the writing of fourth-grade students with writing difficulties. The Journal of Special Education, 42(2), 79-90. doi:10.1177/0022466907310371

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2013). A standardized mean difference effect size for multiple baseline designs across individuals. Research Synthesis Methods, 4(4), 324-341. doi:10.1002/jrsm.1086

Salazar, et al. (2020)

Description

Data from a multiple baseline design conducted by Salazar, Ruiz, Ramírez1, & Cardona-Betancourt (2020). The variables are as follows:

• case. Participant identifier.

• measure. Outcome measure description (AFQ-Y, PTQ-C, or GPQ-C).

• treatment Factor indicating baseline, treatment, post, or follow-up phase.

• time. Measurement occasion.

• outcome. Outcome scores.

Format

A data frame with 324 rows and 5 variables

Source

Salazar, D., Ruiz, F., Ramírez, E., & Cardona-Betancourt, V. (2020). Acceptance and Commitment Therapy Focused on Repetitive Negative Thinking for Child Depression: A Randomized Multiple-Baseline Evaluation. The Psychological Record. doi:10.1007/s40732-019-00362-5

Example from Pustejovsky, Hedges, & Shadish (2014)

Description

Data from a multiple baseline design conducted by Schutte, Malouff, & Brown (2008). Case 4 is excluded because nearly all of these measurements are at the upper extreme of the scale. The variables are as follows:

• case. Participant identifier.

• week. Measurement occasion.

• treatment. Factor indicating baseline or treatment phase.

• fatigue. Fatigue severity scale scores.

Format

A data frame with 136 rows and 4 variables

Source

Schutte, N. S., Malouff, J. M., & Brown, R. F. (2008). Efficacy of an emotion-focused treatment for prolonged fatigue. Behavior Modification, 32(5), 699-713. doi:10.1177/0145445508317133

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

Thiemann & Goldstein (2001)

Description

Data from a multiple baseline across behaviors design conducted by Thiemann & Goldstein (2001). The variables are as follows:

• Study_ID. Study identifier.

• case. Student identifier.

• series. Series identifier.

• outcome. Frequency of coded social communication skills, as measured by a direct observation coding system with 15-second intervals recording for the occurrence of any of the four social measures: contingent responses, securing attention, initiating comments, and initiating requests.

• time. Measurement occasion.

• treatment. Indicator for treatment phase.

• trt_time. Measurement occasion times treatment phase.

• time_c. Measurement occasion centered at the follow-up time.

Format

A data frame with 363 rows and 8 variables

Source

Thiemann, K.S., & Goldstein, H. (2001). Social stories, written text cues, and video feedback: effects on social communication of children with Autism. Journal of Applied Behavior Analysis, 34(4), 425-446. doi:10.1901/jaba.2001.34-425

Thiemann & Goldstein (2004)

Description

Data from a multiple baseline across behaviors design conducted by Thiemann & Goldstein (2004). The variables are as follows:

• Study_ID. Study identifier.

• case. Student identifier.

• series. Series identifier.

• outcome. Frequency of unprompted targeted social communication skills, as measured by a direct observation, paper and pencil coding system during the 10-minute social activity for each behavior for all sessions.

• time. Measurement occasion.

• treatment. Indicator for treatment phase.

• trt_time. Measurement occasion times treatment phase.

• time_c. Measurement occasion centered at the follow-up time.

Format

A data frame with 408 rows and 8 variables

Source

Thiemann, K.S., & Goldstein, H. (2004). Effects of peer training and written text cueing on social communication of school-age children with pervasive developmental disorder. Journal of Speech Language and Hearing Research, 47(1), 126-144. doi:10.1044/1092-4388(2004/012)

Thorne (2005)

Description

Data from an ABAB design conducted by Thorne and Kamps (2008). The variables are as follows:

• case. Participant identifier.

• measure. Outcome measure description (academic engagement or inappropriate verbalizations).

• session. Measurement occasion.

• phase_id. Categorical variable describing the phase of the study design for each case.

• condition Categorical variable describing whether each phase is a baseline (A) phase or intervention (B) phase.

• phase_indicator. Indicator variable equal to 1 during intervention phases.

• outcome. Outcome scores

Format

A data frame with 776 rows and 7 variables

Source

Thorne, S., & Kamps, D. (2008). The effects of a group contingency intervention on academic engagement and problem behavior of at-risk students. Behavior Analysis in Practice, 1(2), 12-18. doi:10.1007/BF03391723

A convenience function for calculating multiple design-comparable effect sizes from a dataset that compiles data from multiple single-case design studies

Description

Calculates standardized mean difference effect sizes for a data set including one or multiple single-case design studies using the same design (treatment reversal, multiple baseline/probe across participants, replicated multiple baseline across behaviors, or clustered multiple baseline across participants).

Usage

batch_calc_BCSMD(
  data,
  design,
  grouping,
  case,
  phase,
  session,
  outcome,
  cluster = NULL,
  series = NULL,
  center = 0,
  round_session = TRUE,
  treatment_name = NULL,
  FE_base = 0,
  RE_base = 0,
  RE_base_2 = NULL,
  FE_trt = 0,
  RE_trt = NULL,
  RE_trt_2 = NULL,
  corStruct = "AR1",
  varStruct = "hom",
  A = NULL,
  B = NULL,
  D = NULL,
  cover = 95,
  bound = 35,
  symmetric = TRUE,
  ...
)

Arguments

Value

A data frame containing the design-comparable effect size estimate, standard error, confidence interval, and other information, for each unique category of grouping variable(s).

Examples

data(Thiemann2001)
data(Thiemann2004)
datThiemann <- rbind(Thiemann2001, Thiemann2004)

# Change-in-levels model with a fixed treatment effect
batch_calc_BCSMD(data = datThiemann,
                 grouping = Study_ID,
                 design = "RMBB",
                 case = case, series = series, phase = treatment,
                 session = time, outcome = outcome,
                 FE_base = 0, RE_base = 0,
                 RE_base_2 = 0, FE_trt = 0)

# Models with linear time trends in baseline and treatment phase,
# random baseline slope at series level, fixed treatment effects
batch_calc_BCSMD(data = datThiemann,
                 grouping = Study_ID,
                 design = "RMBB",
                 case = case, series = series, phase = treatment,
                 session = time, outcome = outcome,
                 FE_base = c(0,1), RE_base = c(0,1),
                 RE_base_2 = 0, FE_trt = c(0,1))

A convenience function for calculating design comparable effect sizes

Description

In one call, 1) clean single-case design data for treatment reversal and multiple baseline designs, 2) fit a multi-level model using restricted maximum likelihood estimation, and 3) estimate a standardized mean difference effect size.

Usage

calc_BCSMD(
  design,
  case,
  phase,
  session,
  outcome,
  cluster = NULL,
  series = NULL,
  center = 0,
  round_session = TRUE,
  treatment_name = NULL,
  FE_base = 0,
  RE_base = 0,
  RE_base_2 = NULL,
  FE_trt = 0,
  RE_trt = NULL,
  RE_trt_2 = NULL,
  corStruct = "AR1",
  varStruct = "hom",
  A = NULL,
  B = NULL,
  D = NULL,
  cover = 95,
  bound = 35,
  symmetric = TRUE,
  summary = TRUE,
  data = NULL,
  ...
)

Arguments

Value

If summary == TRUE, a data frame containing the design-comparable effect size estimate, standard error, confidence interval, and other information. If summary == FALSE, a list containing all elements of a 'g_mlm()' object, plus the fitted 'lme()' model.

Examples

data(Laski)

# Change-in-levels model with fixed treatment effect
calc_BCSMD(design = "MBP",
           case = case, phase = treatment,
           session = time, outcome = outcome,
           FE_base = 0, RE_base = 0, FE_trt = 0,
           data = Laski)

# Model with linear time trends in baseline and treatment phases,
# random baseline slopes, fixed treatment effects
calc_BCSMD(design = "MBP",
           case = case, phase = treatment,
           session = time, outcome = outcome, center = 4,
           FE_base = c(0,1), RE_base = c(0,1), 
           FE_trt = c(0,1),
           data = Laski)


data(Anglesea)
calc_BCSMD(design = "TR",
           case = case, phase = condition,
           session = session, outcome = outcome,
           treatment_name = "treatment",
           FE_base = 0, RE_base = 0, 
           FE_trt = 0,
           data = Anglesea)


data(Thiemann2001)
calc_BCSMD(design = "RMBB",
           case = case, series = series, phase = treatment,
           session = time, outcome = outcome,
           FE_base = 0, RE_base = 0, RE_base_2 = 0, 
           FE_trt = 0,
           data = Thiemann2001)


data(Bryant2018)
calc_BCSMD(design = "CMB",
           cluster = group, case = case, phase = treatment,
           session = session, outcome = outcome, center = 49,
           treatment_name = "treatment",
           FE_base = c(0,1), RE_base = 0, RE_base_2 = 0,
           FE_trt = c(0,1), RE_trt = NULL, RE_trt_2 = NULL,
           data = Bryant2018)

Run simulation comparing REML and HPS estimates

Description

Simulates data from a simple linear mixed effects model, then calculates REML and HPS effect size estimators as described in Pustejovsky, Hedges, & Shadish (2014).

Usage

compare_RML_HPS(iterations, beta, rho, phi, design, m, n, MB = TRUE)

Arguments

Value

A matrix reporting the mean and variance of the effect size estimates and various associated statistics.

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

Examples

compare_RML_HPS(iterations=10, beta = c(0,1,0,0), rho = 0.3, 
                 phi = 0.5, design=design_matrix(m=3,n=8))

Calculate default initial treatment time and follow-up time.

Description

Calculate the default initial treatment time and follow-up time that are used to define and estimate the between-case standardized mean differences for multiple baseline designs and variations.

Usage

default_times(
  design,
  case,
  phase,
  session,
  cluster = NULL,
  series = NULL,
  treatment_name = NULL,
  data = NULL
)

Arguments

Value

A list of time range, default initial treatment time, and default follow-up time.

Note

If treatment_name is left null, it will choose the second level of the phase variable to be the treatment phase.

Examples

data(Laski)
default_times(design = "MBP", 
              case = case, phase = treatment, session = time, 
              data = Laski)

data(Thiemann2001)
default_times(design = "RMBB", 
              case = case, series = series, 
              phase = treatment, session = time, 
              data = Thiemann2001)

data(Bryant2018)
default_times(design = "CMB", 
              cluster = group, case = case, 
              phase = treatment, session = session, 
              data = Bryant2018)
              

Create a design matrix for a single-case design

Description

Create a design matrix containing a linear trend, a treatment effect, and a trend-by-treatment interaction for a single-case design with m cases and n measurement occasions.

Usage

design_matrix(m, n, treat_times = n/2 + 1, center = 0)

Arguments

Value

A design matrix

Examples

design_matrix(3, 16, c(5,9,13))

Calculates HPS effect size

Description

Calculates the HPS effect size estimator based on data from an (AB)^k design, as described in Hedges, Pustejovsky, & Shadish (2012). Note that the data must contain one row per measurement occasion per subject.

Usage

effect_size_ABk(
  outcome,
  treatment,
  id,
  phase,
  time,
  data = NULL,
  phi = NULL,
  rho = NULL
)

Arguments

Value

A list with the following components

Note

If phi or rho is left unspecified (or both), estimates for the nuisance parameters will be calculated.

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2012). A standardized mean difference effect size for single case designs. Research Synthesis Methods, 3, 224-239. doi:10.1002/jrsm.1052

Examples

data(Lambert)
effect_size_ABk(outcome = outcome, treatment = treatment, id = case, 
                phase = phase, time = time, data = Lambert)
   
data(Anglesea)
effect_size_ABk(outcome = outcome, treatment = condition, id = case, 
                phase = phase, time = session, data = Anglesea)

Calculates HPS effect size

Description

Calculates the HPS effect size estimator based on data from a multiple baseline design, as described in Hedges, Pustejovsky, & Shadish (2013). Note that the data must contain one row per measurement occasion per subject.

Usage

effect_size_MB(
  outcome,
  treatment,
  id,
  time,
  data = NULL,
  phi = NULL,
  rho = NULL
)

Arguments

Value

A list with the following components

Note

If phi or rho is left unspecified (or both), estimates for the nuisance parameters will be calculated.

References

Hedges, L. V., Pustejovsky, J. E., & Shadish, W. R. (2013). A standardized mean difference effect size for multiple baseline designs across individuals. Research Synthesis Methods, 4(4), 324-341. doi:10.1002/jrsm.1086

Examples

data(Saddler)
effect_size_MB(outcome = outcome, treatment = treatment, id = case, 
               time = time, data = subset(Saddler, measure=="writing quality"))

data(Laski)
effect_size_MB(outcome = outcome, treatment = treatment, id = case, 
               time = time, data = Laski)

Calculates adjusted REML effect size

Description

Estimates a design-comparable standardized mean difference effect size based on data from a multiple baseline design, using adjusted REML method as described in Pustejovsky, Hedges, & Shadish (2014). Note that the data must contain one row per measurement occasion per case.

Usage

g_REML(
  m_fit,
  p_const,
  r_const,
  X_design = model.matrix(m_fit, data = m_fit$data),
  Z_design = model.matrix(m_fit$modelStruct$reStruct, data = m_fit$data),
  block = nlme::getGroups(m_fit),
  times = attr(m_fit$modelStruct$corStruct, "covariate"),
  returnModel = TRUE
)

Arguments

Value

A list with the following components

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

Examples

data(Laski)
Laski_RML <- lme(fixed = outcome ~ treatment, 
                 random = ~ 1 | case, 
                 correlation = corAR1(0, ~ time | case), 
                 data = Laski)
summary(Laski_RML)
g_REML(Laski_RML, p_const = c(0,1), r_const = c(1,0,1), returnModel=FALSE)

data(Schutte)
Schutte$trt.week <- with(Schutte, unlist(tapply((treatment=="treatment") * week, 
         list(treatment,case), function(x) x - min(x))) + (treatment=="treatment"))
Schutte$week <- Schutte$week - 9
Schutte_RML <- lme(fixed = fatigue ~ week + treatment + trt.week, 
                   random = ~ week | case, 
                   correlation = corAR1(0, ~ week | case), 
                   data = subset(Schutte, case != 4))
summary(Schutte_RML)
Schutte_g <- g_REML(Schutte_RML, p_const = c(0,0,1,7), r_const = c(1,0,1,0,0))
summary(Schutte_g)

Graph Single Case Design Data

Description

Graphs single case design data for treatment reversal and multiple baseline designs.

Usage

graph_SCD(
  design,
  case,
  phase,
  session,
  outcome,
  cluster = NULL,
  series = NULL,
  treatment_name = NULL,
  model_fit = NULL,
  data = NULL,
  newdata = NULL
)

Arguments

Value

A ggplot graph

Note

If treatment_name is left null it will choose the second level of the phase variable to be the treatment phase.

Examples

if (requireNamespace("ggplot2", quietly = TRUE)) {

data(Anglesea)
graph_SCD(design="TR",
          case=case, phase=condition,
          session=session, outcome=outcome,
          treatment_name = "treatment",
          data=Anglesea)

data(BartonArwood)
graph_SCD(design="MBP",
          case=case, phase=condition,
          session=session, outcome=outcome,
          treatment_name = "B",
          data=BartonArwood)

data(Thiemann2001)
graph_SCD(design="RMBB",
          case=case, series = series, phase=treatment,
          session=time, outcome=outcome,
          treatment_name = "treatment",
          data=Thiemann2001)

data(Bryant2018)
graph_SCD(design="CMB",
          cluster=group, case=case, phase=treatment,
          session=session, outcome=outcome,
          treatment_name = "treatment",
          data=Bryant2018)

}

lmeInfo

Description

Functions imported from the lmeInfo package.

• extract_varcomp

• g_mlm

• varcomp_vcov

Calculate phase-pairs for a unique case

Description

Calculate phase-pairs based on phases and session numbering.

Usage

phase_pairs(phase, session = seq_along(phase))

Arguments

Examples

phases <- rep(c("A","B","A","B"), each = 4)
sessions <- 1:length(phases)

phase_pairs(phases, sessions)

phases <- rep(c("A","B","C","A","B","C","D"), each = 4)
phase_pairs(phases)

phases <- rep(c("B","A","C","B","A","B","C","A"), each = 4)
phase_pairs(phases)

Clean Single Case Design Data

Description

Clean single case design data for treatment reversal and multiple baseline designs.

Usage

preprocess_SCD(
  design,
  case,
  phase,
  session,
  outcome,
  cluster = NULL,
  series = NULL,
  center = 0,
  round_session = TRUE,
  treatment_name = NULL,
  data = NULL
)

Arguments

Value

A cleaned SCD dataset that can be used for model fitting and effect size calculation.

Note

If treatment_name is left null it will choose the second level of the phase variable to be the treatment phase.

Examples

data(Laski)
preprocess_SCD(design = "MBP", 
               case = case, phase = treatment,
               session = time, outcome = outcome, 
               center = 4, data = Laski)
               
data(Anglesea) 
preprocess_SCD(design="TR",
               case=case, phase=condition,
               session=session, outcome=outcome,
               treatment_name = "treatment",
               data=Anglesea)
               
data(Thiemann2001)
preprocess_SCD(design = "RMBB", 
               case = case, series = series, phase = treatment, 
               session = time, outcome = outcome, 
               data = Thiemann2001)
          

Calculate session-by-treatment interactions for a unique case

Description

Calculate session-by-treatment interactions based on phases and session numbering.

Usage

session_by_treatment(phase, session, trt_phase)

Arguments

A shiny interface for the scdhlm package

Description

An interactive shiny interface for estimating design-comparable standardized mean difference effect sizes from single-case designs. Estimation methods for multiple baseline and treatment reversal designs are available.

Usage

shine_scd(dataset = NULL, ...)

Arguments

Examples

## Not run: 
shine_scd()
data(Laski)
shine_scd(dataset = Laski)
shine_scd(dataset = "SCD_data.xlsx", sheet = "Laski")
shine_scd(dataset = "Laski.csv") 

## End(Not run)

Simulate data from a fitted g_REML object

Description

Simulates data from the linear mixed effects model used to estimate the specified standardized mean difference effect size. Suitable for parametric bootstrapping.

Usage

## S3 method for class 'g_REML'
simulate(object, nsim = 1, seed = NULL, parallel = FALSE, ...)

Arguments

Value

A matrix with one row per simulation, with columns corresponding to the output of g_REML.

Examples

data(Laski)
Laski_RML <- lme(fixed = outcome ~ treatment,
                 random = ~ 1 | case,
                 correlation = corAR1(0, ~ time | case), 
                 data = Laski)

suppressWarnings(
  Laski_g <- g_REML(Laski_RML, p_const = c(0,1), r_const = c(1,0,1))
)

if (requireNamespace("plyr", quietly = TRUE)) {
  simulate(Laski_g, nsim = 5)
}

Simulate Model MB2 from Pustejovsky, Hedges, & Shadish (2014)

Description

Simulates data from a linear mixed effects model, then calculates REML effect size estimator as described in Pustejovsky, Hedges, & Shadish (2014).

Usage

simulate_MB2(
  iterations,
  beta,
  rho,
  phi,
  tau1_ratio,
  tau_corr,
  design,
  m,
  n,
  MB = TRUE
)

Arguments

Value

A matrix reporting the mean and variance of the effect size estimates and various associated statistics.

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

Examples

set.seed(8)
simulate_MB2(iterations = 5, beta = c(0,1,0,0), rho = 0.4, phi = 0.5, 
             tau1_ratio = 0.5, tau_corr = -0.4, design = design_matrix(m=3, n=8))
             
set.seed(8)
simulate_MB2(iterations = 5, beta = c(0,1,0,0), rho = 0.4, phi = 0.5, 
             tau1_ratio = 0.5, tau_corr = -0.4, m = 3, n = 8, MB = FALSE)
             

Simulate Model MB4 from Pustejovsky, Hedges, & Shadish (2014)

Description

Simulates data from a linear mixed effects model, then calculates REML effect size estimator as described in Pustejovsky, Hedges, & Shadish (2014).

Usage

simulate_MB4(
  iterations,
  beta,
  rho,
  phi,
  tau2_ratio,
  tau_corr,
  p_const,
  r_const,
  design,
  m,
  n,
  MB = TRUE
)

Arguments

Value

A matrix reporting the mean and variance of the effect size estimates and various associated statistics.

References

Pustejovsky, J. E., Hedges, L. V., & Shadish, W. R. (2014). Design-comparable effect sizes in multiple baseline designs: A general modeling framework. Journal of Educational and Behavioral Statistics, 39(4), 211-227. doi:10.3102/1076998614547577

Examples

simulate_MB4(iterations = 5, beta = c(0,1,0,0), rho = 0.8, phi = 0.5, 
             tau2_ratio = 0.5, tau_corr = 0, 
             p_const = c(0,1,0,7), r_const = c(1,0,1,0,0), 
             design = design_matrix(3, 16, treat_times=c(5,9,13), center = 12))
             
simulate_MB4(iterations = 5, beta = c(0,1,0,0), rho = 0.8, phi = 0.5, 
             tau2_ratio = 0.5, tau_corr = 0, m = 6, n = 8)