Projoint Example Data Set 1: Building Conjoint with a Repeated, Flipped Task

Description

A cleaned Qualtrics export from a conjoint study that compares two potential new building developments. Each respondent completed 8 standard tasks as well as a repeated version of the first task (flipped), which can be used to calculate intra-respondent reliability (response instability).

Usage

data(exampleData1)

Format

A data frame with 400 rows and 185 columns. Contains survey responses including demographic information, outcome choices, and conjoint attribute values identified by K-*-* variable names.

Examples

# Load the dataset
data(exampleData1)
# Inspect the first few rows
head(exampleData1)
# Number of rows and columns
dim(exampleData1)
# Display column names (truncated here)
names(exampleData1)[1:10]

Projoint Example Data Set 1: "Labelled Tibble"

Description

A cleaned tibble where each attribute corresponds to a separate column with a descriptive attribute name. The unit of observation is each of two profiles in each task for each respondent.

Usage

data(exampleData1_labelled_tibble)

Format

A tibble with 6,400 rows and 14 columns. Contains survey responses including outcome choices and conjoint attribute values (columns typically named with a 'K-*-*' convention).

Details

This dataset is intended for illustrating reading, reshaping, and analysis workflows in projoint. Column names are compatible with reshape_projoint().

Source

Qualtrics survey on Prolific; see Clayton et al. replication materials.

Examples

# Load the data
data(exampleData1_labelled_tibble)
# Basic inspection (fast and always runnable)
head(exampleData1_labelled_tibble)
dim(exampleData1_labelled_tibble)
# Optional: quick structure peek (names only)
names(exampleData1_labelled_tibble)

Projoint Example Data Set 2: Building Conjoint with a Repeated, Non-Flipped Task

Description

A cleaned Qualtrics export from a conjoint study that compares two potential new building developments. Each respondent completed 8 standard tasks as well as a repeated version of the first task, where the profiles were presented in the same left-right positions (not flipped).

Usage

data(exampleData2)

Format

A data frame with 400 rows and 185 columns. Contains survey responses including demographic information, outcome choices, and conjoint attribute values identified by K-*-* variable names.

Examples

# Load the dataset
data(exampleData2)
# Inspect the first few rows
head(exampleData2)
# Number of rows and columns
dim(exampleData2)
# Display first 10 column names
names(exampleData2)[1:10]

Projoint Example Data Set 3: Building Conjoint with No Repeated Task

Description

A cleaned Qualtrics export from a conjoint study that compares two potential new building developments. Each respondent completed 8 standard tasks only; no repeated tasks are included in this dataset.

Usage

data(exampleData3)

Format

A data frame with 400 rows and 184 columns. Contains survey responses including demographic information, outcome choices, and conjoint attribute values identified by K-*-* variable names.

Examples

# Load the dataset
data(exampleData3)
# Inspect the first few rows
head(exampleData3)
# Number of rows and columns
dim(exampleData3)
# Display first 10 column names
names(exampleData3)[1:10]

Make a projoint_data object from a labelled tibble

Description

Converts a labelled tibble/data frame (one column per attribute) into an object of class projoint_data that downstream functions (e.g., projoint ) can consume. The unit of observation should be each of two profiles in each task for each respondent.

Usage

make_projoint_data(
 .dataframe,
 .attribute_vars,
 .id_var = "id",
 .task_var = "task",
 .profile_var = "profile",
 .selected_var = "selected",
 .selected_repeated_var = NULL,
 .fill = FALSE
)

Arguments

Value

A projoint_data object (a list-like object containing a labels tibble and a data tibble) ready to pass to projoint and related functions.

Examples

# Example: build a projoint_data object from the labelled-tibble example
data(exampleData1_labelled_tibble)
att_cols <- c("School Quality", "Violent Crime Rate (Vs National Rate)",
 "Racial Composition", "Housing Cost",
 "Presidential Vote (2020)", 
 "Total Daily Driving Time for Commuting and Errands",
 "Type of Place")
pj_dat <- make_projoint_data(
 .dataframe = exampleData1_labelled_tibble,
 .attribute_vars = att_cols,
 .id_var = "id",
 .task_var = "task",
 .profile_var = "profile",
 .selected_var = "selected",
 .selected_repeated_var = "selected_repeated",
 .fill = FALSE
)
class(pj_dat)
# [1] "projoint_data"

Organize data for estimation (internal helper)

Description

Prepares tidy inputs for MM/AMCE estimation and IRR handling. Called inside pj_estimate() after reshaping to respondent–task–profile.

Usage

organize_data(
 .dataframe,
 .structure,
 .estimand,
 .remove_ties,
 .att_choose,
 .lev_choose,
 .att_notchoose,
 .lev_notchoose
)

Arguments

Value

A named list with two tibbles:

• $data_for_estimand: rows restricted and reshaped to the requested estimand/QOI. For profile-level, one row per respondent–task–profile where the target level is present (ties optionally removed). For choice-level, one row per respondent–task with paired information for profiles 1 and 2, restricted to the requested attribute-level pair(s).

• $data_for_irr: one row per respondent with columns id, agree (if available), and any detected weight/cluster hints (columns matching "my_weight|my_cluster|^weights?$|^clusters?$").

Example Output: Manually Rearranged Labels

Description

A toy projoint_data object showing what happens after labels have been manually rearranged (e.g., via save_labels / read_labels ).

Usage

data(out1_arranged)

Format

An object of class projoint_data. Contains two elements:

• $labels: reordered attribute–level mapping

• $data: the corresponding profile-level dataset

Details

Useful for illustrating workflows without requiring users to re-run the reordering themselves.

Internal Estimation Function

Description

Core workhorse for computing marginal means (MMs) or AMCEs from a conjoint design, with optional intra-respondent reliability (IRR) correction.

Usage

pj_estimate(
 .data,
 .structure,
 .estimand,
 .att_choose,
 .lev_choose,
 .att_notchoose,
 .lev_notchoose,
 .att_choose_b,
 .lev_choose_b,
 .att_notchoose_b,
 .lev_notchoose_b,
 .se_method,
 .irr,
 .remove_ties,
 .ignore_position,
 .n_sims,
 .n_boot,
 .weights_1,
 .clusters_1,
 .se_type_1,
 .weights_2,
 .clusters_2,
 .se_type_2,
 .auto_cluster = TRUE,
 .seed = NULL
)

Arguments

Details

IRR is clipped to [0.5, 1) (with a tiny epsilon) to avoid boundary issues. For choice-level MMs, ties must be removed (.remove_ties = TRUE). When .seed is supplied, the previous RNG state is restored on exit.

Value

A data frame with rows for the requested estimand(s) and columns:

• estimand: one of "mm_uncorrected", "mm_corrected", "amce_uncorrected", "amce_corrected".

• estimate, se, conf.low, conf.high, tau.

See Also

lm_robust , projoint , projoint_level , projoint_diff

Plot method for projoint_results

Description

Creates publication-ready plots from a projoint_results object produced by projoint . Supports both profile-level and choice-level analyses, with plotting options tailored to each structure.

Usage

## S3 method for class 'projoint_results'
plot(
 x,
 .estimates = "corrected",
 .by_var = FALSE,
 .labels = NULL,
 .base_size = 12,
 .base_family = "",
 .type = c("bar", "pointrange"),
 .show_attribute = TRUE,
 .remove_xaxis = FALSE,
 .xlim = c(0, 1),
 .plot.margin = c(0, 3, 0, 3),
 ...
)

Arguments

Details

For profile-level results, only .by_var, .base_size, and .base_family are relevant. For choice-level results, only .type, .labels, .show_attribute, .remove_xaxis, .xlim, and .plot.margin are relevant. Irrelevant arguments are ignored with a warning.

Value

A ggplot2 object.

See Also

projoint , plot_projoint_profile_level, plot_projoint_choice_level_mm

Examples

data(exampleData1)
# Two base tasks (1 & 2) + repeated of task 1 (last)
dat <- reshape_projoint(
 exampleData1,
 .outcomes = c("choice1", "choice2", "choice1_repeated_flipped")
)
# Build a valid QOI from the labels
att <- unique(dat$labels$attribute_id)[1]
levs <- subset(dat$labels, attribute_id == att)$level_id
lev_names <- sub(".*:", "", levs)
q <- set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose = att,
 .lev_choose = lev_names[2],
 .att_notchoose = att,
 .lev_notchoose = lev_names[1]
)
fit <- projoint(dat, .qoi = q)
# Plot method
plot(fit)

Plot method for projoint_tau

Description

Visualizes the estimated intra-respondent reliability (\tau) produced by the extrapolation method and stored in a projoint_tau object.

Usage

## S3 method for class 'projoint_tau'
plot(x, ...)

Arguments

Value

A ggplot2 object representing the IRR (\tau) visualization. The plot is drawn for its side effect and also returned (invisibly).

Examples

# Estimate tau, then plot:
# dat <- reshape_projoint(exampleData1, .outcomes = c("choice1","choice2"))
# tau_fit <- projoint_tau(dat) # or predict_tau(dat)
# p <- plot(tau_fit) # also returns the ggplot object (invisibly)

Plot choice-level marginal means (MMs) (helper)

Description

Internal helper used by plot.projoint_results to render choice-level marginal means (MMs). Supports a bar chart or a horizontal pointrange layout and optional custom level labels.

Usage

plot_projoint_choice_level_mm(
 x,
 .type = "pointrange",
 .estimates = "corrected",
 .labels = NULL,
 .show_attribute = TRUE,
 .remove_xaxis = FALSE,
 .xlim = c(0, 1),
 .plot.margin = c(top = 1, left = 2, bottom = 1, right = 2),
 ...
)

Arguments

Details

This helper expects that the projoint_results object already contains a single pair of choice-level MMs (i.e., one att_level_choose and one att_level_notchoose). It is called internally by plot.projoint_results when x$structure == "choice_level" and x$estimand == "mm".

When .type = "pointrange", the level labels are placed just outside .xlim to avoid overlap with the confidence interval. If the two levels correspond to different attributes, the attribute title is omitted and a message is emitted.

Value

A ggplot2 object.

See Also

plot.projoint_results for the user-facing plot method.

Examples

data(exampleData1)
dat <- reshape_projoint(exampleData1,
 .outcomes = c("choice1", "choice2", "choice1_repeated_flipped")
)
att <- unique(dat$labels$attribute_id)[1]
levs <- subset(dat$labels, attribute_id == att)$level_id
lev_names <- sub(".*:", "", levs)
q <- set_qoi("choice_level", "mm",
 .att_choose = att, .lev_choose = lev_names[2],
 .att_notchoose = att, .lev_notchoose = lev_names[1]
)
fit <- projoint(dat, .qoi = q)
plot(fit)

Estimate intra-respondent reliability (tau) without a repeated task

Description

Uses the extrapolation method to estimate intra-respondent reliability (IRR, \tau) when your conjoint design does not include an explicit repeated task. The input is a projoint_data object (e.g., produced by reshape_projoint ).

Usage

predict_tau(.data, .title = NULL)

Arguments

Details

The procedure constructs pairs of base tasks within respondent, computes the proportion of identical choices as a function of how many attributes differ between the two tasks, fits a weighted regression of agreement on the number of differing attributes, and extrapolates to zero differences to obtain \hat{\tau}.

Value

A projoint_tau object (list-like) with components:

• $irr: a tibble with columns x (number of differing attributes) and predicted (fitted agreement), including x = 0 which is the estimate of \tau.

• $figure: a ggplot2 object visualizing observed agreement by x and the fitted line with the extrapolated point at x = 0.

See Also

plot.projoint_tau , summary.projoint_tau , reshape_projoint

Examples

# Example workflow:
data(exampleData1)
outcomes <- c(paste0("choice", 1:8), "choice1_repeated_flipped")
# Even if your real study lacks a repeated task, this shows the API:
pj <- reshape_projoint(exampleData1, outcomes, .repeated = TRUE)
tau_fit <- predict_tau(pj, .title = "IRR (tau): ")
# Inspect the extrapolated tau (row where x == 0)
tau_fit$irr[tau_fit$irr$x == 0, ]
# Plot (also available via plot(tau_fit))
print(tau_fit$figure)

Print a projoint_data object

Description

Custom print method for objects of class projoint_data.

Usage

## S3 method for class 'projoint_data'
print(x, ...)

Arguments

Value

No return value, called for its side effect of printing a summary of the projoint_data object to the console.

Examples

data(exampleData1)
dat <- reshape_projoint(
 exampleData1,
 .outcomes = c("choice1", "choice2")
)
print(dat)

Print method for projoint_results

Description

Custom print method for objects of class projoint_results.

Usage

## S3 method for class 'projoint_results'
print(x, ...)

Arguments

Value

No return value, called for its side effect of printing a summary of the projoint_results object to the console.

Examples

data(exampleData1)
dat <- reshape_projoint(
 exampleData1,
 .outcomes = c("choice1", "choice2", "choice1_repeated_flipped")
)
att <- unique(dat$labels$attribute_id)[1]
levs <- subset(dat$labels, attribute_id == att)$level_id
lev_names <- sub(".*:", "", levs)
q <- set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose = att,
 .lev_choose = lev_names[2],
 .att_notchoose = att,
 .lev_notchoose = lev_names[1]
)
fit <- projoint(dat, .qoi = q)
print(fit)

Print method for projoint_tau objects

Description

Custom print method for objects of class projoint_tau, typically created by projoint or related functions.

Usage

## S3 method for class 'projoint_tau'
print(x, ...)

Arguments

Value

No return value; called for its side effect of printing a summary of the estimated intra-respondent reliability (\tau).

Examples

toy_tau <- structure(
 list(irr = data.frame(predicted = 0.413, se = 0.02, n = 200)),
 class = "projoint_tau"
)
print(toy_tau)

Analyze a conjoint dataset with measurement-error correction

Description

Computes marginal means (MMs) or average marginal component effects (AMCEs) with correction for intra-respondent reliability (IRR, \tau). When a repeated task is present, IRR is estimated unless a fixed value is supplied. Results are returned in a structured object ready for plotting and summary.

Usage

projoint(
 .data,
 .qoi = NULL,
 .by_var = NULL,
 .structure = "choice_level",
 .estimand = "mm",
 .se_method = "analytical",
 .irr = NULL,
 .remove_ties = TRUE,
 .ignore_position = NULL,
 .n_sims = NULL,
 .n_boot = NULL,
 .weights_1 = NULL,
 .clusters_1 = NULL,
 .se_type_1 = NULL,
 .weights_2 = NULL,
 .clusters_2 = NULL,
 .se_type_2 = NULL,
 .auto_cluster = TRUE,
 .seed = NULL
)

Arguments

Details

Most users will pass a projoint_data object (from reshape_projoint or make_projoint_data ). Advanced users may specify custom quantities via projoint_qoi ; if provided, its structure and estimand override .structure and .estimand.

Valid se_type_* values depend on clustering:

• Without clusters: "classical", "HC0", "HC1", "HC2", "HC3"

• With clusters: "CR0", "CR1", "CR2", "stata", "none"

If NULL, estimatr defaults are used (HC2 when unclustered; CR2 when clustered).

Value

A projoint_results object (list-like) with components such as:

• $estimand, $structure, $se_method, $irr, $tau

• $labels: attribute/level mapping used in estimation

• $estimates: a data frame of estimates with columns like att_level_choose, att_level_notchoose (if choice-level), estimate, se / std.error, conf.low, conf.high, and an estimand label such as "mm_corrected" or "amce_uncorrected".

This object is suitable for downstream use in plot.projoint_results , summary.projoint_results , and related helpers.

See Also

reshape_projoint , projoint_qoi , plot.projoint_results , summary.projoint_results

Examples

# Prepare example data
data(exampleData1)
outcomes <- c(paste0("choice", 1:8), "choice1_repeated_flipped")
pj <- reshape_projoint(exampleData1, outcomes)
# Choice-level QoI based on pj$labels
att <- unique(pj$labels$attribute_id)[1]
lev_ids <- pj$labels$level_id[pj$labels$attribute_id == att]
lev_names <- sub(".*:", "", lev_ids)
q <- set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose = att,
 .lev_choose = lev_names[2],
 .att_notchoose = att,
 .lev_notchoose = lev_names[1]
 )
 
# Choice-level, marginal means (fast example: fix IRR)
fit_choice <- projoint(
 pj,
 .qoi = q,
 .structure = "choice_level",
 .estimand = "mm",
 .irr = 0.80, # skip IRR estimation for a quick example
 .se_method = "analytical"
)
head(summary(fit_choice))
# Profile-level AMCEs
fit_profile <- projoint(
 pj,
 .structure = "profile_level",
 .estimand = "amce",
 .se_method = "analytical"
)
# Plot using the S3 plot method
p <- plot(fit_profile, .estimates = "both")
print(p)

Create a projoint_data Object

Description

Internal constructor for projoint_data objects. Used by reshape_projoint and make_projoint_data to bundle conjoint survey labels and response data into a consistent structure.

Usage

projoint_data(labels, data)

Arguments

Value

A list of length two with class "projoint_data", containing:

• $labels: attribute–level mapping

• $data: respondent–task–profile dataset

Estimate subgroup differences (internal)

Description

Worker used by projoint to compute subgroup differences (group == 1 minus group == 0) in marginal means (MMs) or average marginal component effects (AMCEs). Supported only for .structure = "profile_level".

Usage

projoint_diff(
 .data,
 .qoi,
 .by_var,
 .structure,
 .estimand,
 .se_method,
 .irr,
 .remove_ties,
 .ignore_position,
 .n_sims,
 .n_boot,
 .weights_1,
 .clusters_1,
 .se_type_1,
 .weights_2,
 .clusters_2,
 .se_type_2,
 .auto_cluster = TRUE,
 .seed = NULL
)

Arguments

Value

A projoint_results object containing subgroup differences with fields:

• estimates: tibble with one row per attribute/level and columns like estimate (group1 minus group0), se, conf.low, conf.high, plus internal columns (estimate_1, estimate_0, se_1, se_0) used in the diff.

• estimand: "mm" or "amce".

• structure: "profile_level".

• se_method: SE method used.

• irr: "Estimated" or "Assumed (<value> )".

• tau: numeric reliability used (average of subgroup taus when estimated).

• remove_ties, ignore_position: flags echoed from inputs.

• se_type_used, cluster_by: details propagated from fitting calls.

• labels, data: design labels and the analysis data (passed through).

See Also

projoint , projoint_level , projoint_results

Estimate profile- or choice-level effects (internal)

Description

Core worker used by projoint to compute marginal means (MMs) or average marginal component effects (AMCEs) under either the profile- or choice-level structure. Handles IRR usage (estimated or fixed) and the requested standard-error method.

Usage

projoint_level(
 .data,
 .qoi,
 .structure,
 .estimand,
 .se_method,
 .irr,
 .remove_ties,
 .ignore_position,
 .n_sims,
 .n_boot,
 .weights_1,
 .clusters_1,
 .se_type_1,
 .weights_2,
 .clusters_2,
 .se_type_2,
 .auto_cluster = TRUE,
 .seed = NULL
)

Arguments

Value

A projoint_results object containing:

• estimates: tibble of estimates (point estimate, SE, CI) with identifier columns (e.g., att_level_*).

• estimand: "mm" or "amce".

• structure: "profile_level" or "choice_level".

• se_method: SE computation method used.

• irr: character noting IRR usage (e.g., "Estimated" or "Assumed (<value> )").

• tau: numeric \tau used to correct measurement error.

• remove_ties, ignore_position: flags echoed from inputs.

• se_type_used, cluster_by: details propagated from fitting calls (if available).

• labels, data: design labels and the analysis data passed through for downstream methods.

See Also

projoint , pj_estimate , organize_data , projoint_results

Create a projoint_qoi Object

Description

Internal constructor for projoint_qoi objects. A projoint_qoi stores the specification of custom quantities of interest (QOIs) for conjoint estimation. Typically called by set_qoi rather than directly.

Usage

projoint_qoi(...)

Arguments

Value

A list of QOI specifications with class "projoint_qoi".

Create a projoint_results Object

Description

Internal constructor for projoint_results objects. A projoint_results stores the outputs of a conjoint analysis, including estimated effects and metadata about the analysis settings. It is normally generated by projoint and not called directly.

Usage

projoint_results(...)

Arguments

Value

A list with class "projoint_results" containing estimates and associated metadata. This object is the standard output of projoint and is accepted by plotting and summary methods.

Read and re-format a Qualtrics CSV (choice text)

Description

Reads a CSV file exported from Qualtrics (with "Use choice text" enabled) and returns a data frame formatted for downstream processing with reshape_projoint .

Usage

read_Qualtrics(.file)

Arguments

Value

A data frame where column names are preserved from the Qualtrics export. The first two rows of Qualtrics metadata are skipped automatically.

See Also

reshape_projoint

Examples

# Write a tiny dummy Qualtrics-style CSV to a temp file
tmp <- tempfile(fileext = ".csv")
readr::write_csv(
 data.frame(Q1 = c("Choice Text", "Choice Text", "A", "B")),
 tmp
)
# Read it back in
df <- read_Qualtrics(tmp)
head(df)

Read and apply a reordered attribute/level mapping

Description

Reads a CSV containing a revised ordering of attributes and levels and applies it to an existing projoint_data object. Typical workflow: first save the current labels to CSV (e.g., with save_labels ), manually reorder rows (and/or the attribute grouping) in the CSV, then call read_labels() to apply.

Usage

read_labels(.data, .filename)

Arguments

Value

A projoint_data object with the same content as .data but with attributes and levels reordered to match the CSV. The returned object contains:

• $labels: a tibble with new attribute_id and level_id reflecting the chosen order

• $data: a tibble whose att* columns have been remapped to the new level_ids

See Also

save_labels , reshape_projoint

Examples

# Create a projoint_data object from the example dataset
data(exampleData1)
outcomes <- c(paste0("choice", 1:8), "choice1_repeated_flipped")
pj <- reshape_projoint(exampleData1, outcomes)
# Write current labels to a temporary CSV, adding an 'order' column
tmp <- tempfile(fileext = ".csv")
pj$labels |>
 dplyr::mutate(order = dplyr::row_number()) |>
 readr::write_csv(tmp)
# (User would reorder rows in 'tmp' manually; we just read it back)
pj_reordered <- read_labels(pj, tmp)
# Inspect the updated label order
head(pj_reordered$labels)

Reshape survey response data for conjoint analysis (single task set)

Description

Takes a wide survey data frame (e.g., from read_Qualtrics ) and reshapes it so that each row corresponds to a single respondent–task–profile. Supports arbitrary ordering of base tasks and a single repeated task per respondent. The repeated base task is inferred from the first base outcome in .outcomes, and the repeated outcome must be the last element of .outcomes.

Usage

reshape_projoint(
 .dataframe,
 .outcomes,
 .choice_labels = c("A", "B"),
 .alphabet = "K",
 .idvar = "ResponseId",
 .repeated = TRUE,
 .flipped = TRUE,
 .covariates = NULL,
 .fill = FALSE
)

Arguments

Details

Scope and assumptions

• One set of conjoint tasks with exactly two profiles per task (profiles 1 and 2).

• For multi-set designs, call reshape_projoint() once per set and bind the results.

Expected input (Qualtrics K-codes)

• Wide columns named K-<task>-<attribute> (attribute names) and K-<task>-<profile>-<attribute> (level names), with <task> in 1..n and <profile> in 1,2.

• Rows with missing K-1-1 are dropped as empty tables (server hiccup safeguard).

Outcome columns (.outcomes)

• List all choice variables in the order asked. If you include a repeated task, its outcome must be the last element.

• For base tasks (all but the last element), the function extracts the base task id by reading the digits in each outcome name (e.g., "choice4", "Q4", "task04" -> task 4).

• The set of base task ids extracted from .outcomes must exactly match the set of task ids present in the K-codes; otherwise an error is thrown.

• The repeated base task is inferred as the digits in the first base outcome (i.e., the first element of .outcomes, excluding the final repeated outcome).

Choice parsing

• The selected profile is parsed from the last character of each outcome string and matched to .choice_labels. Ensure outcomes end with these labels (e.g., "Candidate A"/"Candidate B"). If outcomes are numeric or differently formatted, pre-process or adjust .choice_labels accordingly.

Output

• A projoint_data object with:

  • $labels: map from human-readable attribute/level to stable ids (attribute_id = "att1","att2",..., level_id = "attX:levelY").

  • $data: tibble with one row per id–task–profile, attribute columns (named att*) storing level_id, selected (1 if that profile was chosen; 0 otherwise), agree (1/0/NA for repeated-task agreement after flip logic), and any .covariates. id is coerced to character; attribute columns are factors.

Filling agreement

• If .fill = TRUE, agree is filled within respondent across tasks in task order, propagating the observed repeated-task agreement to all tasks for that respondent. This assumes IRR is respondent-specific and independent of table content.

Diagnostics

• dplyr::count(reshaped$data, task, profile) should show exactly two rows per task.

• If pj_estimate() later reports "No rows match the specified attribute/level", construct QoIs from reshaped$labels (use the exact attX:levelY ids).

Value

A projoint_data object with elements $labels and $data; see Details.

See Also

make_projoint_data , projoint

Examples

# Base tasks asked in numeric order; repeated task corresponds to task 1
data(exampleData1)
outcomes <- c(paste0("choice", 1:8), "choice1_repeated_flipped")
reshaped <- reshape_projoint(exampleData1, outcomes)
dplyr::count(reshaped$data, task, profile) # should be 2 per task

Save attribute and level labels to a CSV file

Description

Saves the attributes and levels (and their order) from a projoint_data object, as generated by reshape_projoint , to a CSV file. This enables manual reordering and later re-import via read_labels .

Usage

save_labels(.data, .filename)

Arguments

Value

No return value, called for side effects (writes a CSV file).

See Also

read_labels , reshape_projoint

Examples

library(projoint)
data(exampleData1)
reshaped <- reshape_projoint(
 exampleData1,
 .outcomes = c(paste0("choice", 1:8), "choice1_repeated_flipped")
)
tmpfile <- tempfile(fileext = ".csv")
save_labels(reshaped, tmpfile)
readLines(tmpfile, n = 5) # show first few lines

Set the quantities of interest (QoIs)

Description

Constructs a quantities-of-interest (QoI) specification for projoint. Use this to request specific estimands—marginal means (MMs) or average marginal component effects (AMCEs)—at either the choice- or profile-level, and to declare which attribute levels are compared (including baselines).

Usage

set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose,
 .lev_choose,
 .att_notchoose = NULL,
 .lev_notchoose = NULL,
 .att_choose_b = NULL,
 .lev_choose_b = NULL,
 .att_notchoose_b = NULL,
 .lev_notchoose_b = NULL
)

Arguments

Value

A projoint_qoi object (list-like) containing fields such as: structure, estimand, attribute_of_interest, levels_of_interest, and their baseline counterparts. This object can be supplied to downstream estimation helpers that accept a QoI spec.

Examples

# Specify a simple choice-level MM comparison for att1 levels:
q_mm <- set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose = "att1",
 .lev_choose = c("att1:lev2"),
 .att_notchoose = "att1",
 .lev_notchoose = c("att1:lev1")
)
str(q_mm)
# Example AMCE with explicit baselines (profile-level):
q_amce <- set_qoi(
 .structure = "profile_level",
 .estimand = "amce",
 .att_choose = "att2",
 .lev_choose = "att2:lev3",
 .att_choose_b = "att2",
 .lev_choose_b = "att2:lev1"
)
str(q_amce)

Summary for projoint_data

Description

Custom summary method for objects of class projoint_data. Prints a brief overview of the main data and attribute-level labels contained in the object.

Usage

## S3 method for class 'projoint_data'
summary(object, ...)

Arguments

Value

No return value, called for its side effect of printing a summary of the projoint_data object to the console.

Examples

data(exampleData1)
dat <- reshape_projoint(
 exampleData1,
 .outcomes = c("choice1", "choice2")
)
summary(dat)

Summary method for projoint_results

Description

Creates a concise tabular summary of a projoint_results object, including the chosen estimand, analysis structure, standard-error settings, and a data frame of estimates.

Usage

## S3 method for class 'projoint_results'
summary(object, ...)

Arguments

Value

A data frame (often a tibble) summarizing the estimated effects. At minimum, it contains the columns produced in object$estimates (e.g., attribute/level identifiers and the point estimate with its standard error and confidence interval in columns such as estimate, std.error, conf.low, conf.high). This table is suitable for further processing or printing.

Examples

 data(exampleData1)
 # Reshape data for two base tasks + repeated (for IRR estimation)
 dat <- reshape_projoint(
 exampleData1,
 .outcomes = c("choice1", "choice2", "choice1_repeated_flipped")
 )
 # Build a valid choice-level QoI
 att <- unique(dat$labels$attribute_id)[1]
 lev_ids <- dat$labels$level_id[dat$labels$attribute_id == att]
 lev_names <- sub(".*:", "", lev_ids)
 q <- set_qoi(
 .structure = "choice_level",
 .estimand = "mm",
 .att_choose = att,
 .lev_choose = lev_names[2],
 .att_notchoose = att,
 .lev_notchoose = lev_names[1]
 )
 # Fit model
 fit <- projoint(dat, .qoi = q)
 # Get the tabular summary of estimates
 tab <- summary(fit)
 head(tab)

Summary method for projoint_tau objects

Description

Custom summary method for objects of class projoint_tau, typically created by projoint or related functions. Summarizes intra-respondent reliability (IRR) estimates.

Usage

## S3 method for class 'projoint_tau'
summary(object, ...)

Arguments

Value

A tibble (data frame) showing IRR estimates, typically by the number of differing attributes, as stored in object$irr.

Examples

toy_tau <- structure(
 list(irr = data.frame(predicted = 0.413, se = 0.02, n = 200)),
 class = "projoint_tau"
)
summary(toy_tau)