R Implementation of item level descriptive summary

Description

The des_summary function provides descriptive statistics for numerical and categorical variables in the study data.
Depending of the type of data, the function provides the appropriate measures of central tendency (i.e., mean, median, and mode); measures of dispersion (i.e., standard deviation, interquartile range, mean absolute deviation, range of values, and coefficient of variation); information on the number of categories and their frequency, on the shape of the distribution (skewness and kurtosis), and on missing data. It also provides plots to give an overview of the data distribution.

Usage and arguments

des_summary(
  resp_vars = NULL,
  study_data = sd1,
  label_col = LABEL,
  meta_data = md1
)

The function has the following arguments:

• resp_vars: optional, a character specifying the measurement variables of interest. If missing, all variables from the study_data are assessed;
• study_data: mandatory, the data frame containing the measurements;
• meta_data: mandatory, the data frame containing the item-level metadata. If this refers to missing tables, then these have to be existing as files or URLs or be loaded before using prep_load_workbook_like_file(), prep_load_folder_with_metadata(), or prep_get_data_frame();
• label_col: optional, the column in the metadata data frame containing the labels of all the variables in the study data.

Example output

To illustrate the output, we use the example synthetic data and metadata that are bundled with the dataquieR package. See the introductory tutorial for instructions on importing these files into R, as well as details on their structure and contents.

prep_load_workbook_like_file("meta_data_v2")
sd1 <- prep_get_data_frame("study_data")

des_sum <- des_summary(
  study_data = sd1,
  label_col = LABEL
)

The function generates 2 outputs SummaryData and SummaryTable, that are exactly the same in this case, but used differently in the creation of a report.

Output 1: Summary data frame

The summary data frame is called using des_sum$SummaryData:

Either as an interactive data.tables table:

DT::datatable(des_sum$SummaryData, escape = FALSE)

Or as a kable:

Variables	Labels	STUDY_SEGMENT	Mean	Median	Mode	SD	IQR (Quartiles)	MAD	Range (Min - Max)	CV	Skewness	Kurtosis	No. categories/Freq. table	Valid	Missing	Graph
CENTER_0	Examination center CENTER_0 v00000 [nominal, integer]	STUDY			Berlin								No. unique values (incl. NA): 5 ‘Berlin’ ‘Leipzig’ ‘Munich’ ‘Hamburg’ ‘Cologne’ 632 602 597 592 577	3000 100.000%	0 0.000%
DBP_0	Diastolic blood pressure DBP_0 v00005 [ratio, float]	PHYS_EXAM	81.29	81	80	9.2142	12 (Q1 = 75 | Q3 = 87)	8.896	57 (54 - 111)	11.335	0.0805 (0.0447)	-0.5562		2544 84.800%	456 15.200%
GLOBAL_HEALTH_VAS_0	Self-reported global health GLOBAL_HEALTH_VAS_0 v00006 [ratio, float]	PHYS_EXAM	5.027	5	3.2	2.9184	5.075 (Q1 = 2.5 | Q3 = 7.575)	3.706	10 (0 - 10)	58.0511	-0.0015 (0.0447)	-1.4368		2618 87.267%	382 12.733%
ASTHMA_0	Known asthma ASTHMA_0 v00007 [nominal, integer]	PHYS_EXAM			no								No. unique values (incl. NA): 3 ‘no’ ‘yes’ 2117 524	2641 88.033%	359 11.967%
VO2_CAPCAT_0	Aerobic capacity category VO2_CAPCAT_0 v00008 [ordinal, string]	PHYS_EXAM		good	excellent		3 (Q1 = excellent | Q3 = restricted)		(excellent - pathological)				No. unique values (incl. NA): 6 ‘excellent’ ‘good’ ‘moderate’ ‘restricted’ ‘pathological’ 784 647 500 380 284	0 0.000%	405 13.500%
ARM_CIRC_0	Upper arm circumference ARM_CIRC_0 v00009 [ratio, float]	PHYS_EXAM	25.033	25	24	3.9576	6 (Q1 = 22 | Q3 = 28)	4.448	27 (11 - 38)	15.8097	-0.0237 (0.0447)	-0.3594		2657 88.567%	343 11.433%
ARM_CIRC_DISC_0	Upper arm circumference cat ARM_CIRC_DISC_0 v00109 [ordinal, integer]	PHYS_EXAM		(20,30]	(20,30]		0 (Q1 = (20,30] | Q3 = (20,30])		((-Inf,20] - (30, Inf])				No. unique values (incl. NA): 4 ‘(20,30]’ ‘(-Inf,20]’ ‘(30, Inf]’ 2071 344 218	2633 87.767%	367 12.233%
ARM_CUFF_0	Upper arm circumference device ARM_CUFF_0 v00010 [ordinal, integer]	PHYS_EXAM		(20,30]	(20,30]		0 (Q1 = (20,30] | Q3 = (20,30])		((-Inf,20] - (30, Inf])				No. unique values (incl. NA): 4 ‘(20,30]’ ‘(-Inf,20]’ ‘(30, Inf]’ 2015 351 257	2623 87.433%	377 12.567%
USR_VO2_0	Aerobic capacity examiner USR_VO2_0 v00011 [nominal, string]	PHYS_EXAM			USR_321								No. unique values (incl. NA): 16 ‘USR_321’ ‘USR_590’ ‘USR_213’ ‘USR_592’ ‘USR_211’ Others 449 301 223 223 216 1370	0 0.000%	218 7.267%
USR_BP_0	Blood pressure examiner USR_BP_0 v00012 [nominal, string]	PHYS_EXAM			USR_301								No. unique values (incl. NA): 16 ‘USR_301’ ‘USR_243’ ‘USR_537’ ‘USR_542’ ‘USR_123’ Others 448 347 319 208 201 1252	0 0.000%	225 7.500%
EXAM_DT_0	Examination date and time EXAM_DT_0 v00013 [interval, datetime]	PHYS_EXAM	2018-07-02 10:09:59 UTC	2018-07-05 19:45:30 UTC	2018-03-21 20:44:05 UTC 2018-04-12 05:25:04 UTC and other 3 dates	9064795 secs	15347279 secs (Q1 = 2018-04-03 04:24:05 UTC | Q3 = 2018-09-27 19:32:04 UTC)	11520753 secs	364 days (2018-01-01 UTC - 2018-12-31 UTC)					2940 98.000%	60 2.000%
PART_PHYS_EXAM	Physical exam consent PART_PHYS_EXAM v20000 [nominal, integer]	PHYS_EXAM			yes								No. unique values (incl. NA): 2 ‘yes’ 2940	2940 98.000%	60 2.000%
CRP_0	C-reactive protein CRP_0 v00014 [ratio, float]	LAB	2.888	2.587	0.16	1.8053	2.27 (Q1 = 1.608 | Q3 = 3.878)	1.637	11.894 (0.118 - 12.012)	62.5065	0.8966 (0.0447)	0.9983		2699 89.967%	301 10.033%
BSG_0	Erythrocyte sedimentation rate BSG_0 v00015 [ratio, float]	LAB	14.857	11	10	12.1348	14 (Q1 = 6 | Q3 = 20)	10.378	96 (0 - 96)	81.6771	1.3774 (0.0447)	2.678		2686 89.533%	314 10.467%
DEV_NO_0	Device ID DEV_NO_0 v00016 [nominal, integer]	LAB			2								No. unique values (incl. NA): 6 ‘2’ ‘3’ ‘1’ ‘4’ ‘5’ 661 626 593 412 400	2692 89.733%	308 10.267%
LAB_DT_0	Lab analysis date and time LAB_DT_0 v00017 [interval, datetime]	LAB	2018-07-02 12:00:36 UTC	2018-07-05 21:50:00 UTC	2018-01-01 02:00:00 UTC 2018-01-10 15:55:28 UTC and other 59 dates	9064818 secs	15342119 secs (Q1 = 2018-04-03 06:21:05 UTC | Q3 = 2018-09-27 20:03:04 UTC)	192027.4 mins	364.0007 days (2018-01-01 02:00:00 UTC - 2018-12-31 02:01:00 UTC)					2940 98.000%	60 2.000%
PART_LAB	Lab analysis consent PART_LAB v30000 [nominal, integer]	LAB			yes								No. unique values (incl. NA): 2 ‘yes’ 2940	2940 98.000%	60 2.000%
EDUCATION_0	Highest educational level B/L EDUCATION_0 v00018 [ordinal, integer]	INTERVIEW		uppersecond	uppersecond		2 (Q1 = secondary | Q3 = postsecond)		(pre-primary - secondtertiary)				No. unique values (incl. NA): 8 ‘uppersecond’ ‘secondary’ ‘postsecond’ ‘primary’ ‘tertiary’ Others 568 557 432 376 258 281	2472 82.400%	528 17.600%
EDUCATION_1	Highest educational level F/U EDUCATION_1 v01018 [ordinal, integer]	INTERVIEW		uppersecond	uppersecond		2 (Q1 = secondary | Q3 = postsecond)		(pre-primary - secondtertiary)				No. unique values (incl. NA): 8 ‘uppersecond’ ‘secondary’ ‘postsecond’ ‘primary’ ‘tertiary’ Others 562 541 428 360 260 271	2422 80.733%	578 19.267%
FAM_STAT_0	Marital status FAM_STAT_0 v00019 [nominal, integer]	INTERVIEW			1								No. unique values (incl. NA): 5 ‘1’ ‘2’ ‘0’ ‘3’ 845 721 588 235	2389 79.633%	611 20.367%
MARRIED_0	Currently married MARRIED_0 v00020 [nominal, integer]	INTERVIEW			no								No. unique values (incl. NA): 3 ‘no’ ‘yes’ 2005 361	2366 78.867%	634 21.133%
SEX_0	Sex B/L SEX_0 v00002 [nominal, integer]	STUDY			females								No. unique values (incl. NA): 3 ‘females’ ‘males’ 1478 1462	2940 98.000%	60 2.000%
N_CHILD_0	Number of children N_CHILD_0 v00021 [ratio, integer]	INTERVIEW	2.499	2	2	1.5297	2 (Q1 = 1 | Q3 = 3)	1.483	9 (0 - 9)	61.2097	0.4907 (0.0447)	-0.2911		2336 77.867%	664 22.133%
EATING_PREFS_0	Eating preferences EATING_PREFS_0 v00022 [nominal, integer]	INTERVIEW			none								No. unique values (incl. NA): 4 ‘none’ ‘vegetarian’ ‘vegan’ 1366 712 250	2328 77.600%	672 22.400%
MEAT_CONS_0	Meat consumption MEAT_CONS_0 v00023 [ordinal, integer]	INTERVIEW		1-2d a week	never		2 (Q1 = never | Q3 = 3-4d a week)		(never - daily)				No. unique values (incl. NA): 6 ‘never’ ‘3-4d a week’ ‘1-2d a week’ ‘5-6d a week’ ‘daily’ 946 475 382 326 173	2302 76.733%	698 23.267%
SMOKING_0	Current smoker SMOKING_0 v00024 [nominal, integer]	INTERVIEW			no								No. unique values (incl. NA): 3 ‘no’ ‘yes’ 1585 707	2292 76.400%	708 23.600%
SMOKE_SHOP_0	Purchasing tobacco products SMOKE_SHOP_0 v00025 [ordinal, integer]	INTERVIEW		3-4d a week	3-4d a week		2 (Q1 = 1-2d a week | Q3 = 5-6d a week)		(never - daily)				No. unique values (incl. NA): 6 ‘3-4d a week’ ‘5-6d a week’ ‘daily’ ‘1-2d a week’ ‘never’ 176 169 154 150 133	782 26.067%	2218 73.933%
N_INJURIES_0	Number of injuries N_INJURIES_0 v00026 [ratio, integer]	INTERVIEW	4.588	4	4	2.4221	3 (Q1 = 3 | Q3 = 6)	2.965	14 (0 - 14)	52.7921	0.4743 (0.0447)	-0.5724		2199 73.300%	801 26.700%
N_BIRTH_0	Number of births N_BIRTH_0 v00027 [ratio, integer]	INTERVIEW	3.458	3	3	1.7707	2 (Q1 = 2 | Q3 = 4)	1.483	12 (0 - 12)	51.2115	0.2105 (0.0447)	-1.7027		1099 36.633%	1901 63.367%
INCOME_GROUP_0	Income group INCOME_GROUP_0 v00028 [ordinal, integer]	INTERVIEW		[30-50k)	[30-50k)		2 (Q1 = [10-30k) | Q3 = [50-70k))		(below 10k - above 90k)				No. unique values (incl. NA): 7 ‘[30-50k)’ ‘[10-30k)’ ‘[50-70k)’ ‘below 10k’ ‘[70-90k)’ Others 622 584 395 294 205 74	2174 72.467%	826 27.533%
PREGNANT_0	Currently pregnant PREGNANT_0 v00029 [nominal, integer]	INTERVIEW			no								No. unique values (incl. NA): 3 ‘no’ ‘yes’ 996 69	1065 35.500%	1935 64.500%
MEDICATION_0	Medication use MEDICATION_0 v00030 [nominal, integer]	INTERVIEW			yes								No. unique values (incl. NA): 2 ‘yes’ 292	292 9.733%	2708 90.267%
AGE_0	Age B/L AGE_0 v00003 [ratio, integer]	STUDY	49.914	50	51	4.4232	6 (Q1 = 47 | Q3 = 53)	4.448	30 (33 - 63)	8.8616	-0.0367 (0.0447)	-0.1761		2940 98.000%	60 2.000%
N_ATC_CODES_0	Number of ATC codes N_ATC_CODES_0 v00031 [ratio, integer]	INTERVIEW	2.262	1	0	2.7257	3 (Q1 = 0 | Q3 = 3)	1.483	22 (0 - 22)	120.4786	1.4057 (0.0447)	3.0554		2058 68.600%	942 31.400%
USR_SOCDEM_0	Sociodemographics examiner USR_SOCDEM_0 v00032 [nominal, string]	INTERVIEW			USR_321								No. unique values (incl. NA): 16 ‘USR_321’ ‘USR_247’ ‘USR_520’ ‘USR_120’ ‘USR_125’ Others 380 297 290 172 159 988	0 0.000%	714 23.800%
INT_DT_0	Interview date and time INT_DT_0 v00033 [interval, datetime]	INTERVIEW	2018-07-02 12:40:34 UTC	2018-07-05 22:27:30 UTC	2018-08-15 22:55:34 UTC 2018-08-24 01:57:42 UTC	9064782 secs	15348014 secs (Q1 = 2018-04-03 06:44:50 UTC | Q3 = 2018-09-27 22:05:04 UTC)	11522577 secs	363.9972 days (2018-01-01 02:24:00 UTC - 2018-12-31 02:20:00 UTC)					2940 98.000%	60 2.000%
PART_INTERVIEW	Interview consent PART_INTERVIEW v40000 [nominal, integer]	INTERVIEW			yes								No. unique values (incl. NA): 3 ‘yes’ ‘no’ 2924 16	2940 98.000%	60 2.000%
ITEM_1_0	Item 1 ITEM_1_0 v00034 [ratio, integer]	QUESTIONNAIRE	3.037	3	3	1.764	2 (Q1 = 2 | Q3 = 4)	1.483	9 (0 - 9)	58.0849	0.4138 (0.0447)	-0.5894		2248 74.933%	752 25.067%
ITEM_2_0	Item 2 ITEM_2_0 v00035 [ratio, integer]	QUESTIONNAIRE	2.988	3	2	1.7008	2 (Q1 = 2 | Q3 = 4)	1.483	10 (0 - 10)	56.9277	0.4079 (0.0447)	-0.648		2197 73.233%	803 26.767%
ITEM_3_0	Item 3 ITEM_3_0 v00036 [ratio, integer]	QUESTIONNAIRE	3.014	3	3	1.7175	2 (Q1 = 2 | Q3 = 4)	1.483	10 (0 - 10)	56.9898	0.4105 (0.0447)	-0.6058		2184 72.800%	816 27.200%
ITEM_4_0	Item 4 ITEM_4_0 v00037 [ratio, integer]	QUESTIONNAIRE	3	3	3	1.7214	2 (Q1 = 2 | Q3 = 4)	1.483	10 (0 - 10)	57.3701	0.4347 (0.0447)	-0.5563		2143 71.433%	857 28.567%
ITEM_5_0	Item 5 ITEM_5_0 v00038 [ratio, integer]	QUESTIONNAIRE	6.021	6	6	2.3738	4 (Q1 = 4 | Q3 = 8)	2.965	10 (0 - 10)	39.4237	-0.1831 (0.0447)	-1.3516		2074 69.133%	926 30.867%
ITEM_6_0	Item 6 ITEM_6_0 v00039 [ratio, integer]	QUESTIONNAIRE	5.948	6	6	2.3706	4 (Q1 = 4 | Q3 = 8)	2.965	10 (0 - 10)	39.8567	-0.1194 (0.0447)	-1.4139		2048 68.267%	952 31.733%
AGE_GROUP_0	Age group B/L AGE_GROUP_0 v00103 [ordinal, string]	STUDY		50-59	50-59								No. unique values (incl. NA): 5 ‘50-59’ ‘40-49’ ‘60-69’ ‘30-39’ 1554 1322 39 25	0 0.000%	60 2.000%
ITEM_7_0	Item 7 ITEM_7_0 v00040 [ratio, integer]	QUESTIONNAIRE	6.037	6	6	2.4007	4 (Q1 = 4 | Q3 = 8)	2.965	10 (0 - 10)	39.7687	-0.1739 (0.0447)	-1.3982		2068 68.933%	932 31.067%
ITEM_8_0	Item 8 ITEM_8_0 v00041 [ratio, integer]	QUESTIONNAIRE	5.895	6	6	2.3974	4 (Q1 = 4 | Q3 = 8)	2.965	10 (0 - 10)	40.6699	-0.0971 (0.0447)	-1.4874		2013 67.100%	987 32.900%
QUEST_DT_0	Questionnaire date and time QUEST_DT_0 v00042 [interval, datetime]	QUESTIONNAIRE	2018-08-04 14:59:47 UTC	2018-07-30 08:21:46 UTC	2018-03-17 01:21:08 UTC 2018-04-07 19:44:05 UTC and other 4 dates	12050776 secs	16542372 secs (Q1 = 2018-04-21 09:39:33 UTC | Q3 = 2018-10-29 20:45:45 UTC)	12215994 secs	1032.387 days (2018-01-11 04:26:00 UTC - 2020-11-08 13:43:28 UTC)					2931 97.700%	69 2.300%
PART_QUESTIONNAIRE	Questionnaire consent PART_QUESTIONNAIRE v50000 [nominal, integer]	QUESTIONNAIRE			yes								No. unique values (incl. NA): 3 ‘yes’ ‘no’ 2864 76	2940 98.000%	60 2.000%
AGE_1	Age F/U AGE_1 v01003 [ratio, integer]	STUDY	49.872	50	51	4.4291	6 (Q1 = 47 | Q3 = 53)	4.448	30 (33 - 63)	8.8808	-0.0315 (0.0447)	-0.1855		2940 98.000%	60 2.000%
SEX_1	Sex F/U SEX_1 v01002 [nominal, integer]	STUDY			females								No. unique values (incl. NA): 3 ‘females’ ‘males’ 1472 1468	2940 98.000%	60 2.000%
PART_STUDY	Study consent PART_STUDY v10000 [nominal, integer]	STUDY			yes								No. unique values (incl. NA): 2 ‘yes’ 2940	2940 98.000%	60 2.000%
SBP_0	Systolic blood pressure SBP_0 v00004 [ratio, float]	PHYS_EXAM	126.516	127	130	9.613	13 (Q1 = 120 | Q3 = 133)	8.896	63 (97 - 160)	7.5982	0.0636 (0.0447)	-0.5639		2561 85.367%	439 14.633%

Interpretation

Algorithm of the implementation

1. From the metadata, determine the scale level and the data type of each variable.
2. For variables of scale level nominal, calculate the mode, and the number of categories and their frequency.
3. For variables of scale level ordinal, calculate the median, the mode, the range of values (min - max), the interquartile range (1st and 3rd quartiles), and the number of categories and their frequency
4. For variables of scale level interval, calculate the mean, the median, the mode, the standard deviation (SD), the MAD, the range of values (min - max), the interquartile range (1st and 3rd quartiles), the coefficient of variation. In addition, only for data types integer or float, calculate the skewness and its standard error, and the kurtosis.
5. For variables of scale level ratio, calculate the mean, the median, the mode, the standard deviation (SD), the MAD, the range of values (min - max), the interquartile range (1st and 3rd quartiles), the coefficient of variation. In addition, only for data types integer or float, calculate the skewness and its standard error, and the kurtosis.
6. Generate a data frame containing the summary for each variable.

Concept relations

• Data quality Indicator Unexpected location

Impressum – Legal Notice       Contact and Funding       CC BY-SA 4.0