EMAeval Description

Noah Kraus 11/03/2020

EMAeval R Markdown

The R package EMAeval contains functions created to help researchers identify careless responses as well as responders ine EMA data. An example dataset is included in the package to help the user better understand the uses of each function. This dataset “EMAeval_Data” is used in the example code below.

There are 4 “participants,” each having 50 assessments. Each was asked 8 questions per assessment.

EMAeval_Data Example

	ID	StartDate	EndDate	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8
1	1001	2020-01-02 10:01:23	2020-01-02 10:02:07	61	9	46	39	5	64	74	32
56	1002	2020-01-07 10:01:16	2020-01-07 10:01:24	45	50	49	48	54	55	48	52
111	1003	2020-01-12 10:00:46	2020-01-12 10:12:08	72	99	25	68	23	42	21	25
172	1004	2020-01-23 09:58:22	2020-01-23 09:58:38	15	5	12	30	73	83	36	6

Functions

Below are the functions in the R package EMAeval. The functions are:

• flagging_df
• flagging_plots
• TPI_cutoff
• SD_cutoff
• Perc_Mode_cutoff
• Combined_cutoff
• Combined_cutoff_percent

Each section will be dedicated to a particular function, giving an example of the usage with the EMAeval_Data and showing the output.

flagging_df

*This function creates a dataframe that reports Time to Complete (TTC), Time per Item (TPI), Item Standard Deviation (SD), and Longstring. If the Longstring returns NA, then there was no longstring response because all item responses were different. The partial results of the flagging_df function are based from the EMAeval_Data Example above. *

flaggingDF <- flagging_df(EMAeval_Data, 
                          ttc.colnames = c("StartDate", "EndDate"), 
                          item.colnames = colnames(EMAeval_Data[,4:11]))

Flagging Dataframe Example

	TTC	TPI	SD	Longstring
1	44	5.50	25.21	NA
56	8	1.00	3.36	48
111	682	85.25	29.31	25
172	16	2.00	30.22	NA

flagging_plots

This function creates a histograms of each of the calculations reported in the flagging_df function. This can be used to help users identify the cutoff values for TPI and SD.

flagging_plots(EMAeval_Data, 
               ttc.colnames = c("StartDate", "EndDate"), 
               item.colnames = colnames(EMAeval_Data[,4:11]),
               number.items = 8)

Note: the Longstring histogram has a much smaller scale for the Count. This is due to the lack of longstring values because many assessments do not have a Longstring value because all item responses differ.

TPI_cutoff

This function creates a dataframe of ID and data indices in which the assessment met the cutoff criterion for Time per Item. The user inputs their own cutoff for TPI. The user can also specify what type of comparison they would like to complete with the cutoff value using condition =… If responses to all items are mandatory, then the following response should be included:

mandatory.response = TRUE

Below is the code for the function.

TPI_cutoff(EMAeval_Data, 
           cutoff = 1, 
           condition = "<=",  
           ttc.colnames = c("StartDate", "EndDate"),  
           number.items = 8,
           mandatory.response = TRUE, 
           item.colnames = colnames(EMAeval_Data[,4:11]), 
           ID.colname = "ID")

Assessments Flagged by TPI Cutoff

ID	Index_of_Flagged_Assessment
1002	51
1002	52
1002	53
1002	54
1002	55
1002	56
1002	57
1002	58
1002	62
1002	63
1002	65
1002	67
1002	68
1002	69
1002	70
1002	71
1002	72
1002	75
1002	76
1002	78
1002	79
1002	80
1002	81
1002	82
1002	83
1002	87
1002	90
1002	91
1002	92
1002	93
1002	96
1002	97
1002	98
1002	99
1003	116

SD_cutoff

This function creates a dataframe of ID and data indices in which the assessment met the cutoff criterion for Item Score Standard Deviation. The user inputs their own cutoff for SD. The user can also specify what type of comparison they would like to complete with the cutoff value using condition =…

Below is the code for the function.

SD_cutoff(EMAeval_Data, 
          cutoff = 5,  
          condition = "<=", 
          item.colnames = colnames(EMAeval_Data[,4:11]), 
          ID.colname = "ID")

Assessments Flagged by SD Cutoff

ID	Index_of_Flagged_Assessment
1002	51
1002	52
1002	53
1002	54
1002	55
1002	56
1002	57
1002	58
1002	59
1002	60
1002	61
1002	62
1002	63
1002	64
1002	65
1002	66
1002	67
1002	68
1002	69
1002	70
1002	71
1002	72
1002	73
1002	74
1002	75
1002	76
1002	77
1002	78
1002	79
1002	80
1002	81
1002	82
1002	83
1002	84
1002	85
1002	86
1002	87
1002	88
1002	89
1002	90
1002	91
1002	92
1002	93
1002	94
1002	95
1003	142

Perc_Mode_cutoff

This function creates a dataframe of ID and data indices in which the assessment met the cutoff criterion for the Percent of Items at Mode. The user inputs their own cutoff for Percent of Items at Mode. The user can also specify what type of comparison they would like to complete with the cutoff value using condition =…

Below is the code for the function.

Perc_Mode_cutoff(EMAeval_Data, 
                 cutoff = 0.7,  
                 condition = ">=", 
                 item.colnames = colnames(EMAeval_Data[,4:11]), 
                 ID.colname = "ID")

Assessments Flagged by PErcent of Items at Mode Cutoff

ID	Index_of_Flagged_Assessment
1002	100
1003	138
1003	140
1003	142
1003	150

Combined_cutoff

This function creates a dataframe of ID and data indices in which the assessment met the cutoff criteria for Time per Item OR Item Score Standard Deviation OR Percent of Items at Mode. The user inputs their own cutoff for TPI, SD and Percent of Items at Mode. The user can also specify what type of comparison they would like to complete with each cutoff value using either SD.condition =… and TPI.condition=… and Perc.Mode.condition=… Users can also specify the logical component for the criteria, specifying with Combined.logic = … If responses to all items are mandatory, then the following response should be included:

mandatory.response = TRUE

Below is the code for the function.

Combined_cutoff(EMAeval_Data, 
                SD.cutoff = 5, SD.condition = "<=", 
                TPI.cutoff = 1, TPI.condition = "<=", 
                Perc.Mode.cutoff = 0.7, Perc.Mode.condition = ">=", 
                Combined.logic = "or", 
                ttc.colnames = c("StartDate", "EndDate"), 
                number.items = 8, mandatory.response = TRUE, 
                item.colnames = colnames(EMAeval_Data[,4:11]), 
                ID.colname = "ID")

Assessments Flagged by TPI, SD or Percent of Items at Mode Cutoffs

ID	Index_of_Flagged_Assessment
1002	51
1002	52
1002	53
1002	54
1002	55
1002	56
1002	57
1002	58
1002	59
1002	60
1002	61
1002	62
1002	63
1002	64
1002	65
1002	66
1002	67
1002	68
1002	69
1002	70
1002	71
1002	72
1002	73
1002	74
1002	75
1002	76
1002	77
1002	78
1002	79
1002	80
1002	81
1002	82
1002	83
1002	84
1002	85
1002	86
1002	87
1002	88
1002	89
1002	90
1002	91
1002	92
1002	93
1002	94
1002	95
1002	96
1002	97
1002	98
1002	99
1003	116
1003	142

Combined_cutoff_percent

This function creates a dataframe of ID and percent of responses in which assessments met the cutoff criteria for Time per Item OR Item Score Standard Deviation OR Percent of Items at Mode. The user inputs their own cutoff for TPI, SD and Percent of Items at Mode. The user can also specify what type of comparison they would like to complete with each cutoff value using either SD.condition =… and TPI.condition=… and Perc.Mode.condition=… Users can also specify the logical component for the criteria, specifying with Combined.logic = … If responses to all items are mandatory, then the following response should be included:

mandatory.response = TRUE

Below is the code for the function.

Combined_cutoff_percent(EMAeval_Data, 
                        SD.cutoff = 5, SD.condition = "<=", 
                        TPI.cutoff = 1, TPI.condition = "<=",  
                        Perc.Mode.cutoff = 0.7, Perc.Mode.condition = ">=", 
                        Combined.logic = "or",
                        ttc.colnames = c("StartDate", "EndDate"),  
                        number.items = 8, mandatory.response = TRUE, 
                        item.colnames = colnames(EMAeval_Data[,4:11]), 
                        ID.colname = "ID")

Percentage of Assessments Flagged by TPI, SD or Percent of Items at Mode Cutoffs

ID	Percent_Flagged
1002	100
1003	10